I guess Linus doesn't care about the environment or any of the other issues surrounding A.I. I'm not anti-AI just for the sake of being anti-AI. I have a huge problem with AI bcz of it's environmental impacts, intellectual property violations, job impacts, attempts to replace all employees with AI, the general dumbing down of society and worse off the internet is bcz of AI, the surveillance state being built around it. I could go on, but you get the point. I am a staunch, no AI person bcz of all those reasons and a hundred more.
This isn't an age thing, or a, I am getting older and Im stubborn thing. AI is literally ruining the planet and everything on it, and this guy thinks we should just accept that.
I say fuck off with that noise, I don't have e to accept anything.
It's perfectly ok to use a new tool to do your work faster, more efficiently, and with greater precision if these improvements are justified by the cost of the tool. But if that tool is a coal-fired, water hoarding, ophan crushing, billionaire-owned kleptobot, your results need to be better than pure gold to be worth the true cost of using it.
I've thought since the beginning of all of this that the biggest mistake was calling all of these LLMs and other tools that they smashed together "AI". It's misleading and ruins the reputation of all of the tools. If someone had made a coding tool that ONLY did what "AI" does with coding, no one would give a shit. But now because an LLM spelled "Strawberry" with too many "r's" well that means it's all slop and it's all shit.
As for the people throwing shade for ethics, EVERYONE breaks their ethics for consumption and EVERYONE has their justifications. I'm not saying that environmental issues aren't important, they very much are, but giving others shit for their choices when half of your clothing and your phone partly relies on slave labor and stripping of the environment weakens your attempt. It's easy to sit on a high horse and judge others for their use of AI. It's much more difficult to understand the nuances of why they do. Just like you would want someone else to try to understand the nuances of why your shoes cost $150 and the 9 year old who made them was only paid 2 cents.
The point that I'm trying to make here, is that you can complain about AI without shitting on others.
"the kind of expectations i have is that people do their best and i expect people to understand what they are doing."
my immediate response to some idiot thinking i accuse him of cheating because i voiced concerns over his methodologies: "look, if you want to drink mercury while writing code and dont see any problems with that, fine. go ahead. waste away your health. but at least have the decency of sitting under the venting hood and stop making it everyone elses problem."
I got banned from the Fuck AI community for saying Linus isn’t stupid. Not that he’s right on AI, just that he’s not stupid.
Indeed, the atmosphere is toxic.
We’re on a site where people happily discuss their arr stack or transpiring the PS5 system, but bemoan AI is killing artists and coding.
There are real ethical issues, job impacts, environmental issues, property rights issues (real and intellectual), privacy, a stack of cards economy etc.
But there is also utility, and frankly the genie is out of the bottle. For better or worse.
Catching kernel bugs in Linux with AI should be seen as a good because they’re not being exploited offensively by private actors. AI review that catches the most annoying 80% of errors or issues saves human time, while not contributing slop to the kernel.
There is a lot to discuss if you show up in good faith.
Also, Linus is an asshole. Why would anyone be surprised he reacted like an asshole? His point on his expectations is actually clear and sensible to me. He probably could have said just that, but he’s an asshole, and acting how you would expect after a bunch of people are yelling at him.
Nuance is a hard concept to get in most Lemmy comment threads. It's obvious it's just a tool that uses an insane amount of resources that can do some good in certain applications. The problem stems from it being abused for evil and stupid things.
Can AI be used to help flag cancerous growths in x rays and help save lives? Absolutely! Can it also be used to track every human being on the planet for surveillance purposes while its proponents build massive biome-destroying data centers to act as the infrastructure? It sure can!
All that said, I still don't feel confident about rolling in AI code to open source projects when bots aren't accountable to their contributions. They should require a huge amount of oversight which I doubt is happening at the moment.
I want to expand on your first thought and just say that in my experience the internet as a whole is not a great place for nuanced takes, and I've noticed that becoming mirrored in reality over the last decade. Perhaps its been happening for longer.
There's so much going on, and so many actors trying to influence the masses. I don't blame anyone for having different conclusions to my own, there's an almost untrackable amount of sources to draw information (and disinformation) from. I just hope we can get to a point together where we can (again?) discuss nuance without ceding to emotion.
On AI contributed code I’m always confused, if I put MY name on it and push it I am stating it meets my standards. If I allow a Claude contributed commit into my code, I am stating Claude passed my standards.
If someone is contributing and it’s not up to standards, why is it getting merged? If it’s a maintainer and they’re merging crap with review or standards then you couldn’t have actually trusted them pre-LLMs.
Ever since the first iteration of GitHub Copilot it was always your responsibility to make sure contributions to your project were correct. The beauty of open source is that when a project jumps off a cliff you can fork it or make your own.
Partially what is good about Linus is historically Linux has said no to crap, and the Linux team expend considerable effort to remove crap they realized was crap too late.
From my layperson understanding, the true program is not the lines of code, it's the model that lives in the programmer's brain. The code is just an extension of that.
If it needs debugging, it is the mental model that helps. If you need to add a feature, same thing. If a few years go by and no one remembers how the code works, rebuilding that mental model is the first step in making progress.
I suspect that it used to be that if someone contributed good code to the kernel project, it meant that they had a sufficient mental model. But with AI, it can seem like they do.
Maybe if you're Linus your understanding of the kernel is so strong it would take years to atrophy, if at all. I don't know. But developers in general?
Too many times in my work life, people have told me one thing, but it turns out that AI invented that and they didn't really know what they were talking about. On things that I'm an expert in, I can catch it. But if I'm not an expert, I don't realize the mistake until a consequence occurs.
Allowing AI is basically making it so the maintainers have to be complete experts in everything they accept. Maybe that was already the case, but that is my layperson understanding of it.
This is correct. And yes, we expect everyone that contributes to have this understanding. If you don't have it, welcome to ask. Misunderstandings that happen despite your effort will be caught by the steward.
Just don't engage with any community that starts with "fuck" if you have any kind of nuance for the subject matter.
It's not a great time. If you don't get banned, you'll get dogpiled with angry and idiotic comments anyway, by people that have decided their own facts already.
Vague criticisms, fallacious reasoning, highly emotive language, little response to attempts at discussion (though some might engage in disingenuous discussions of shifting goalposts here and there to discourage particularly vocal engagees from future participation). So I'd say that Linus is just the latest convenient scapegoat in an Anti-AI 'trend' that will have little impact on the rate of AI adoption in the long run.
Sucks that I waited until now to try to form a measured opinion on AI because now all the discussions on the internet are hot garbage.
Discussions on the internet are hot garbage in part because thanks to LLMs you can't know what's real anymore. Of course there were bots before, but any random idiot can pay OpenAI for a subscription now and spam anywhere they want.
There's nothing Marxist in that article. Only fallacies from rich white man ranting about people pushing back against fascism.
People say AI to mean LLM first because main stream western, fascist, media pushed the term by inviting exclusively non-expert and CEOs to talk about the subject. People are pushing back against a product that is mean to make workers dull and replace them by stealing their works even by capitalist standards -> exploiting their works, appropriating it to themselves without any compensation nor permission.
The article tried to paint the discussion as anti-llm would be anti technology, which is not Marxist in any way and not relevant. Weirdly enough fascist guy like the one on this article don't talk about cryptocurrency as something we should chill for. Isn't it a tool too?
Current LLM chill is the opposite of dialectical materialism: what is the concrete analysis here? LLM outputed nothing of value for years despite the hundred of billions invested. Worst it actually participate to global warming, bourgeoisie attempt to mass control, getting rid of workers and inflating computer costs. So the real consequences for people are losses only.
And for what? To get to have unskilled rich white people who can afford the awfully hight price of LLM to boats about "I made this" while it is actually made from steal and empoverish people and enriching tech lord's.
So no, thanks, and trying to paint fallacies to Marxism isn't going to help either.
EDIT: replace the "stealing" argument as pointed out -> https://lemmy.ml/post/54102413/28384201 , this can be confusing with the anti-piracy propaganda vocabulary.
Actually let me analyze again your comment here: I thought you were right at first but thinking about it not really.
First let me address : I didn't know about RIAA, this kind of paternalistic way of addressing people is not good in any way. We aren't all USA-centric (like you?).
Then about the "steal" argument:
it is a vocabulary used for century by Marxist organizations about the value taken by the bourgeoisie from the proletariat work.
the argument of "copying is OK" is missing the point : any FOSS dev/organization I know would gladly give their work to you, individual. But not to big tech to commercialize it and make extreme amount of money, as well as steering conversation towards fascism in the process. While also having an amazingly big impact on the environment.
"both have it afterwards" is also a fallacy : any work done on computer can be seen like this : if your an accountant and you boss copy your accounting sheets and do not pay you, it is not stealing? You both have the file, for you it won't make you any money, he will. Big tech taking FOSS code for their profit is not piracy as of individual accessing cultural product they cannot afford in the first place. And if you make profit from piracy this is a whole other discussion I would gladly here your thoughts about it.
I checked your history and you seems to genuinely interact but you're sentence makes no real argument (more like a joke).
But maybe you haven't read the rest of the conversation so here some points:
You can and are encourage to download and use free software. Same with comments.
You are not allowed to monetized free software (hence the GPL v2 vs v3 debate). To be honest also with my comments.
The term stealing could indeed be confused with the modern usage of fascists who want to fight against piracy but it was used for a good century by communist to talk about capitalists stealing surplus value made by workers.
Private property is theft. "Intellectual property" is a form of private property. It exclusively benefits the bourgeois.
Regarding your third bullet point: Try your example again, but this time the boss doesn't copy the file. It's still labor theft. The copying is irrelevant.
Are you saying I am right then? I feel you think of something but it is not clearly getting to me here.
it's labor theft -> hence the term steal I was using
copying is irrelevant -> so I was right? the labor theft of big tech in order to sell LLM is wrong on many level and we should fight against it.
Do you at least see that, to me, what you just said is validating my points?
Also if anyone else (not to you @m532): "AI" (in reality chatbot or generative AI) should be made by the people for the people, and FOSS ideology (can we say Librism in English?) have always, and still mostly is, against the closed-source (close dataset to be precise here) nature of current big-tech AI. This was already a problem 8 years ago when the space started to get filled with "scientific" paper that didn't disclose code/data (sometime just code but still insufficient) hence couldn't be reproduce while sometimes having wild claims. We shouldn't give so much credit, attention, money to those fascists in charge of selling as much LLM as possible, whatever human, economic and environment costs.
Cory Doctorow's "Reverse Centaurs Guide to Life After AI" has a short section that pertains to this. He talks about how someone like Linus essentially doesn't have a boss and has untold resources for whatever "AI" "tools" he finds useful.
He isn't ever forced to use the wrong tool for the job or use a tool that simply doesn't do what his boss thinks it does. He doesn't have to worry about getting fired or his coworkers getting fired because his boss was sold a story about reducing labor costs with this cool new "tool". He simply uses what's useful to him with the probably millions of dollars of compute made available to him by the tech giants.
95% of people don't have that luxury. They have a boss itching to get rid of them or 10x their workload with the help of the "tool" that Linus is so happily and defiantly praising on behalf of the monopolists.
I don't see how this means he's wrong though, if people are hating AI as a proxy for hating their abusive relationship with their employer, that doesn't seem like a strong argument that everyone who isn't in that situation must also hate AI.
It doesn't mean he's wrong! it means he's out of touch and doing the propaganda work of capital for free much like many other self-employed/business owning/petit-borgeois people who think they're on the side of capital.
It also means that his praise for and indeed positive experience with this "tool" is largely irrelevant to the broader social/economic discourse surrounding so called AI.
Blind hatred for AI as a technology is inherantly reactionary, but that doesn't mean that the implementation of the technology today isn't problematic in many ways.
Yeah I notice he doesn't address the costs, economic and environmental, of LLM proliferation. Way to ignore a large part of the argument while calling anyone opposed to your position stupid.
I hope he gets to eat those words one day, but I doubt he would ever admit he was wrong.
Then the picture would be significantly changed if the cost of running useful LLMs goes down to what it costs to acquire a reasonable computer to work on the Linux kernel, no? That is if we could run useful LLMs for kernel programming on our kernel programming computers, and we own those LLMs. In such a world, capital would have no stronger leverage over us than it already does by selling us the computers we need to program. Does this argument make sense?
It also means that his praise for and indeed positive experience with this “tool” is largely irrelevant to the broader social/economic discourse surrounding so called AI.
Yeah, but tbf all of the statements he's making about it are in the context of organizing linux development and the most effective way to carry that out, and just happen to be public and so end up being widely republished and incorporated into discussions with a broader topic than that.
Shush! They are not supposed to know that the AI–Anti-Ai fight is fabricated bullshit with both sides being equally stupid (like people arguing based on ideological zeal always are...). And that it's just a show so that the majority is occupied with watching the mud fight so they don't think about the actual problem.
What's the point of this argument though? Sure it's real for every profession and every tool but what's the agentic choice we ought to pursue here? Whine against AI or actually lead society with policy? As "ban llm" is not a valid policy that would ever work so the only viable path is regulation and growing of culture where AI is used for good, no?
Embrace AI, work towards open weight/free software and protect worker rights. Society wins at the bleeding edge not at conservstism.
The answer is certainly not to embrace "AI" in it's current forms or allow for big tech monopolists to continue their wasteful and destructive infrastructure buildout. The oligarchy isn't going to regulate itself to worker self-determination either. Nothing short of the death of capitalism will do.
Ok, let me know how that goes. While you whine online and have no impact on reality we will embrace "AI" and drain billionaire funds for distillation, open weights and building free software.
You genuinely don't see the effective differences of thse two strategies? Actually doing something vs having your finger crossed some magical boogeyman called "capitalism" will collapse?
I think there's a broader question here from an evolutionary perspective. Humans are fundamentally tool users, and technology has been central to our development as a species. It's possible that this evolutionary niche we've occupied will eventually reach its limits and contribute to our own downfall, but I'm not convinced AI is uniquely responsible. It's part of a much longer trajectory of technological progress.
We've evolved to solve problems and use tools to make our lives easier. That doesn't mean technological progress is inevitable, but these tendencies, combined with competition, make it difficult to imagine us ever really stopping. Even if we stopped today, who's to say another generation of humans 10,000 years from now wouldn't continue down the same path?
I don't think this is exclusively about capitalism. Capitalism certainly incentivizes growth and reducing labor costs, but competition, greed, and the pursuit of technological advantages have existed under other economic systems too. That doesn't diminish the legitimate concerns about how AI is being used against workers.
I also wonder whether our short lifespans make us particularly bad at considering the long-term consequences of our innovations. Who is going to sacrifice the opportunities of their own generation to prevent a possible catastrophe centuries from now, especially when there's no guarantee it would stop technological progress?
I have a degree in psychology and particularly enjoyed evolutionary psychology, so I think it's worth asking whether the deeper problem is our own nature, and whether our capacity for collective restraint can keep pace with our ability to innovate.
I know this raises more questions than it answers, but we've faced civilization-changing technologies before, from the printing press and electricity to nuclear weapons. AI won't be the last. Perhaps the more important question is how we learn to manage these transformations when our tendency to create them seems so deeply rooted in who we are.
This argument is conflating the bad boss with the misattribution of the tool. You could do that for any tool throughout history. But I still fail to see how you go from "bad boss is bad because they make us misattribute tool" to "good boss is bad because he's not wallowing with me about how bad boss is bad".
As anyone who has known about him for more than a couple of years knows, Linus has a pretty solid track record of calling out stupidity. It's just virtually always stupidity that affects the kernel. If he had to stop and tell bosses at other companies how to operate, Linux never would have made it out of the 90s.
You can predict with certainty what a compilers output would be. Linus is conflating the fact that most people don't understand how a compiler converts higher level programming language into machine code with the reality that no one can actually predict and audit what an LLM puts out.
He is saying that you have to understand your contribution, so whatever the LLM spits out, you need to thoroughly review and double-check that it makes sense. So, in particular, he's not trying to rate these tools, he's merely saying make sure the contribution is objectively good, then he doesn't care how you got there.
Predicting what a compiler outputs is quiet hard, especially if you enable optimization flags or get near any undefined behavior. But I think you're right to say that a compiler is more predictable than any LLM.
I don't get the point why you should be unable to audit the output of an LLM? If its used as a tool like described the user should have the knowlage to audit any output
The other practical side of this is enforceability.
Lets say Linux banned all use of VSCode in contributions, this second. And any Microsoft tools. No exceptions. And it makes sense, because there is an evil blatantly anti-linux entity behind it... What are people gonna do?
Keep using it anyway, and just not disclose it.
This is why linux doesn't tell people what tools they can and can't use. Practically, they can't police what tools people use, just what they contribute. Just like (to a reasonable degree) they don't try to police people who contributes to the kernel, but instead judge them by their contributions. Other contributors don't have to like those tools or contributors, but that is how a project like this must operate.
It relies on legality, not honor. E.G. with GPL, or code with a proprietary license. It's the threat of losing their job, getting sued or whatever that discourages plagarism.
Ah I get your point now. Yes, ultimately both policies would rely on honor.
I'd argue plagiarism is different, though. If a codebase stipulates 'don't plagiarize,' that's... common morality. Not many contributors are going to argue 'well, I want to plagiarize.' They wouldn't advertise it in any project they contribute to.
Dictating what tooling to use is different. That's going to fall under 'personal choice' for more contributors. And they may feel compelled to hide tool use for that particular project, when they would otherwise show its use publicly in others.
In other words, forbidding tools like, projects runnings be telling contributors to forgo something they're used to using public. That's going to trigger more lack-of-disclosure and rule-breaking from devs that otherwise wouldn't do that.
It’s a pretty reductive take, that treats tools as politically and behaviorally inert. The work on understanding how cultural artifacts use us just as much as (or maybe more than) we use them goes back several centuries.
But Linus is just an engineer, after all. There’s no reason to treat him as some high-minded philosopher (and he’d probably be the first to tell you that). And for as long as the history of thought into the politics of artifacts may be, the history of engineers focusing only on solving the problem in front of them and not thinking about “where the rockets come down” is about as long.
I'm kind of gobsmacked by this. I've heard younger developers compare LLM with compiler and chalked it up to ignorance, but I don't get what this is.
Can I write code in assembly? Yes. Of course. It's true that I never do it, but if there was a rule that you needed to write some assembly now and then to certify as a kernel developer I don't see how it would be any kind of impediment.
Comparing more directly: if you hand me a compiled binary and the source code and ask "why is the bit at address 0xXXXXXXXXX a 0?" I can, with 100% certainty, answer that. Probably in less than a day of effort. If I ask "why did the LLM output "for" as the XXXXth word of its utterance, can Linus say why?
And "can you write a good commit message without an editor?" Uhhh? At least when I was in school, this was something people did for fun/messing around. I still do cat > somefile when I need to put some quick text together. In fact, Claude will sometimes write a Git commit by cat-ing into a file--which it's doing because there's so many online examples of this. Fucking weird things for Linus to pick. Makes me think his brain's rotted.
I'm finding the allure of AI is particularly strong for people who have risen to positions of power where they had to stop doing the work themselves. Instead, they had to work through other people to get bigger projects done.
The allure is that you can just say what you want done, and it does it. If it's done it "wrong" you can just tell it it's wrong and not have to worry about arguing the technical differences. As such you feel like you get closer to what you wanted, rather than somebody else's take on what you wanted.
The things is, when you are removed from the code, you can only work at a high level. That person you asked to do something is considering 10 other low level concerns that you didn't consider. I think Linus's recent patch series is a good example. He saw a potential optimisation, did a proof of concept but missed a race condition. He's now handed it over to the person that normally works in that area and knows the details. Can the optimisation actually work? Who knows?
AI is not a tool. Its an addiction. Maybe the first dozen times (more likely: never) you use it "as a tool". Very quickly, you depend on it. You need it to refactor. Then you need it to debug. Then you need it to write. Next thing you know, you spend 0 time writing code (the fun part) and 100% of your time as a reviewer (the worst part). Suddenly it isnt fun. A few fools even get TWO AI MODELS, one to write, the other to debug.
Its a virus. I cannot tell you how many 300+ line commits ive rejected because they straight up dont do what the person said they do or also do some extremely asinine 2nd thing that breaks other pieces of code (which they would know if theyd actually used it).
A tool requires input. A hammer can never do anything i dont explicitly tell it to do. A good word editor does exactly what i want it to do. A good compiler does exactly what it says it will do. AI does not. Ai is a glorified autocomplete, and i dont want my code to be automatically generated. Its asking for thousands of issues down the line. If AI was any good, Windows would be a usable platform by now.
Dont get me wrong; i know that it is sometimes used as a tool by people, and meticulously checked over, but more often than not, it isnt.
I feel about ai like i feel about cars. 90% of the idiots out there drive really poorly, but that doesn't mean responsible people are smart to never drive a car. Same with AI, overwhelming majority of people are using AI incorrectly, i.e. recklessly dangerously, but that doesn't mean responsible people should never use it.
AI is much more than the quality of the code. It's also about stealing the knowledge, overusing Earth's resources like water and power, pollution, killing consumer computers through RAM and memory shortage in general - the very same computers that run Linux -, making average people more ignorant, risking another gigantic bubble burst and so on.
And as a cherry on the top, comparing AI to compilers. Riiiight.
Have you actually looked into the real environmental impact of AI is compared to, say, eating meat, driving a car, or taking a yearly plane flight? It's a rounding error. Have you looked into the numbers? Compared the environmental impact of your AI use to the environmental impact of taking a yearly vacation on a plane?
The RAM shortage is real, but very temporary. Check the news today! Chip stocks like Sk Hynix are way down because people are worried about over-production. The fact that a new technology caused a temporary shortage in a resource just isn't even a coherent argument against the technology. There's discussion to be had here about whether we should as a civilization reserve some resources for other industries. That's a discussion we could have. But the idea that AI caused a temporary shortage in RAM, so that makes it bad is... like, really ridiculous? You must see that that doesn't even make sense as an argument?
These are not arguments that hold up to scrutiny. You should consider whether you accepted these arguments because you looked into the details and they actually make sense, or if you accepted them because they're telling you what you want to believe.
You're free to hate GenAI all you like as far as I'm concerned. Just be honest about your reasons, and don't pretend it's about the environment if you don't care enough to even look at the numbers.
Sure, but surely the data centers illegally running gas turbines in Memphis and Vineland, New Jersey are causing problems that wouldn't exist without the current hype cycle.
Do you think that's a reasonable reason to hate gen AI?
There are absolutely issues around the rollout of AI. I haven't seen anyone disagree with that. That's just not a good argument for being opposed to the technology. That might be a reasonable reason to hate xAI or Microsoft (and I do!). It's definitely a reasonable reason to have concerns about how we're doing this. You should be concerned.
It's not a reasonable reason to hate the technology itself.
Basically you are saying that me buying and throwing a plastic bottle in to the nature is fine, since one bottle is nothing. Also I'm sure that (new) data centers run on few AAA batteries.
Listen, I've got nothing against you hating AI. There's lots of valid reasons to be concerned, especially with the implications for the future of jobs and work. The way corporations are adopting AI is legitimately concerning.
Just don't pretend it's about the environment when it's not. That isn't just an unconvincing argument, it hurts the environmental movement generally.
It's the plastic straws all over again. You understand that the stupid straws debate did the environment no favors right? Fighting AI as an environmental issue is much worse because the AI is coming, and you will lose that fight. If you're arguing for more controls about water and power use, I'm with you. If you're arguing for a carbon tax, I will sing your praises. If you're trying to fight the idea of AI itself, and pretending that it's about the environment, that's where I have a problem.
I confess, I’m guilty. I’ve used AI to diagnose random issues with my little personal blog (please don’t hate me, I’m just a man without a degree). That said, I understood the simple code well enough to see what I was getting wrong and comprehend the AI generated suggestion, so it wasn’t all that different than looking things up on a wiki and implementing it. The primary difference, for me, was that I had already spent two weeks digging deeper and deeper for my answers, only finding outdated or incorrect information, and Claude delivered the correct answer in about 10 seconds.
In my opinion, the reason it was a good idea to implement the AI generated code was that I understood it. There was a human brain looking at that result and going “aha! I see what I was missing!” If I hadn’t understood what I was looking at, it could have been doing anything, and I wouldn’t have felt comfortable implementing it. If I hadn’t at least tried to the old fashioned way first, I wouldn’t have used it. But after weeks of hitting dead ends and with the knowledge that there was an extremely powerful tool that could give me my answer in seconds, it began to feel foolish to continue punishing myself with wasted time.
In Linus’ case, I think it is safe to say he understands enough about Linux to interpret the AI suggestions and be a safeguard against slop. I’m not too concerned about his contributions to the kernel if he’s using AI to think outside the box or diagnose a complicated problem.
But! There is a major risk factor when people start implementing code that they don’t understand, or when it is a crutch for their total lack of comprehension. Vibe coding and using a diagnostic tool are different things. I hope Linus gives some real thought to the realistic limits that need to be in place for people who aren’t knowledgeable enough to filter AI suggested code through their own brains, because it wouldn’t take long to destroy a reliable system like Linux if they incorporate AI generated code with no experienced human in the drivers seat.
Apologies to those who just fundamentally hate AI and think I’m a fool for using it, I just wanted to share my experience as an unschooled computer nerd who finally gave in and tried it out a couple of times.
Educated computer nerd here. AI has fundamentally changed how you do CS in industry, it's completely unavoidable at this point and not using it in some capacity is doing yourself a massive disservice. It's able to catch edge cases we'd never have caught otherwise. The important part is indeed understanding what it's doing and keeping contributions organized/concise.
That doesn't make it any less scary from an ethics and model oversight perspective, but from the perspective of the project it is fundamentally just another tool. Linus appears to feel this way too. Frankly I'd be more concerned if there weren't people using AI on Linux, especially to validate kernel level security.
I mean, I use it because I'm forced to, but I don't feel much in productivity gains (not using agents though). Sometimes it's very, very handy, but most of the time it's like talking to the stupidest rubber ducky, and the frustration it gives me is not worth that imo. Others might make a different calculation, but ultimately if search engines didn't enshitify themselves into obsoletion, and LLMs didn't produce the worst, most verbose, and sometimes objectively wrong documentation, I'd not see an upside at all. But here we are.
AI has fundamentally changed how you do CS in industry, it’s completely unavoidable at this point and not using it in some capacity is doing yourself a massive disservice
If it was in truth inevitable then there would be no need to constantly shriek "It is inevitable!"
There is also evidence that deferring critical thinking to an LLM causes cognitive atrophy.
Conversations on this subject will continue to stagnate and miss the important points as long as people continue to act like the discussion of ethics is the same as the discussion of efficacy. People who disagree with AI on a principle of its impacts being negative will often call it “a lying machine which is regularly wrong and offers no utility”.
May I suggest that people with criticisms of the ethics and impact of AI take a charitable position towards the utility offered by AI. Accept something along the lines of a positive account given by an experienced software engineer - it’s useful, it presents risks but also it presents new ways to get certain work done well in much less time etc.
Even if you do not believe this, the point is it is a much more ambiguous argument than the ethical argument or the concerns over the impact of AI on the species, environment or society. You weaken those arguments by putting them along side poorly justified, somewhat falsifiable claims that AI just doesn’t work.
I think he's kinda right in that competent engineers can write good code using LLMs. I don't know whether they're actually more productive that way though. The problem lies in the fact that incompetent engineers can write a lot of bad, but good-looking code using LLMs, which might be a pretty big burden on the maintainers.
Also, comparing LLMs to compilers doesn't really work in my opinion. Compilers follow strict rules and translate one kind of logic into a different kind of logic, while LLMs turn natural language and vague notions into some kind of logic. Writing code is telling the computer precisely what it should do and prompting is vaguely telling it what end result you want and those are two fundamentally different things.
the thing is - no other tool (currently ^yall_cant_read^) in our industry, and others, out there involves as much active destruction in both the working class and environment in its use.
that’s it. that’s the bit.
for those not understanding: the word “in” above implies in current use not that i am comparing AI’s effects with the invention of knitting machines or electricity.
Are your clothes machine knitted? Do you know you're destroying the jobs of people who hand knitted clothes? And let's not forget the art form of hand knitting that is now being insulted by people who buy machine knitted clothes without a care. And let's not forget about the environmental impact of knitting machines that require electricity while hand knitting doesn't!
The same could be said about so many things.
There isn't really a fundumental difference thar I can see between AI and other tools, at least that I can see.
Students are using it to do their homework for them and learning nothing; at least with CliffsNotes you got the gist of what you were supposed to be reading. We shouldn't want a dumber society.
GenAI also makes it substantially easier to post propaganda and spam. Bots on the internet were already common but now trying to find genuine art or conversations is like pulling up steel from ships that sank before 1945 because everything since then has been irradiated.
The US's dumbfuck government almost started a war with China earlier this year thanks to someone using GenAI. And yeah everyone knows they're a bunch of fucking idiots but that doesn't change the fact that it wouldn't have happened if LLMs didn't exist.
AI in school is an interesting issue, we can't discount the fact that there are benefits to using AI in school as well. It needs to be up to countries to regulate usage and educate children on it's use. Socialist countries seem to be doing a pretty good job at this.
The US’s dumbfuck government almost started a war with China
"sources say" hmm I wonder what those sources are, I'm sure the CNN would never extrapolate information out of nowhere to spread sinophobia and call for AI regulation (which is exactly what AI companies want at the moment.)
Calling something just a tool is whitewashing who’s making it and promoting it and what it’s used for. Stop saying that AI is just a tool and it only matters how it is used
I'm not sure how calling AI a tool is whitewashing. Of course tools don't exist outside of their material conditions, however, AI has many benefits, it's just that capitalist economies will always use tools for the detriment of workers. Socialist countries have economies that are orientated to benefit the working class. They recognise the problems that arise from AI and work to fix them while using its benefits as much as possible to liberate workers rather than burden them.
I don't think AI is the most destructive at all. AI is responsible for around 1% of carbon emissions (do your own research if you want a proper source). There are so many things, such as cars that emit way more than this.
As for job loss there are so many things that have eliminated so many more jobs. Tractors, knitting machines, production lines, automation in factories etc. No one says we should stop using tractors so we can have more jobs for farmers.
also: i do knit, lol. just a coincidence in example.
That's funny, I've got a half knitted scarf somewhere but I've never gotten into it properly.
Now do it for the planned capacity of every AI data center currently planned or under construction or waiting for compute manufacturing by the big tech monopolists, and the trillions of dollars of labor that they're telling the world they'll replace whether the product is capable of it or not.
The bot doesn't have to be able to do your job to displace your labor. It only has to convince your boss that it can. Right now the big tech monopolists are actively convincing the entire world's bosses that they can replace your labor. This doesn't have to be true to be a labor issue, and it certainly doesn't have to be true to be a devastating economic bubble.
Now do it for the planned capacity of every AI data center currently planned or under construction or waiting for compute manufacturing by the big tech monopolists, and the trillions of dollars of labor that they’re telling the world they’ll replace whether the product is capable of it or not.
No you. My statistic still stands, if you would like to provide a different one please do. Maybe it will be 2%.
Your second paragraph is shifting the blame away from capitalism and towards AI. Economic bubbles and mass layoffs are features of capitalism, this is a pattern that has been repeating since capitalism was born. If not AI it would be something else.
I don't believe they'll finish. If you do believe they'll finish and succeed in ripping communities off as planned, then I think you're also predicting that this stuff really works and soon.
I think the real problem is that sometimes things like this come around that are disruptive to the workforce, faster than the workforce can handle the influx of new unemployment.
It will in theory sort itself out, but at the moment, were in a transition period.
That doesn't mean its bad, but it can mean that our governments aren't responding in ways that support the worker in the transition which is their fault, not the technologies fault.
Edit: essentially disruption will always keep happening, its how its handled thats the important part.
essentially disruption will always keep happening, its how its handled thats the important part.
I do agree. We can look at countries with different economic and political systems to see which ones handle them best. Countries with socialist economic systems are able to use AI for the benefit of workers. While companies with capitalist economic systems use AI at the expense of workers.
I beg you people to just once honestly put the supposed resource use of AI into any kind of perspective. What do you think, how many prompts worth of AI did your choice to eat a beef burger instead of a bean burger cost the world? How many have you eaten, have all the AI haters eaten, across their life time?
That has been SO much worse than the negatives we've seen with AI.
If anyone is interested, that is a horrible story socially, industrially, environmentally, etc.
Definitely worth reading about.
Well, the issue is that people want to turn off AI, but not these other things. It's like, only the newest bad thing counts. The bad things I like (like meat every day) are actually fine, somehow...
Explain to me how Photoshop is comparably destructive as AI is lol. Also the cotton gin replaces slave labour primarily. Idk if that's what you want to put up as an example here.
I don't have a personal or professional workflow that would benefit from AI. I have not been impressed with any generated art or writing (knowingly, so far). I pretty much only know it for the enshittification of search results. It might be great at coding, I'll never know because in my first 49 years of life I've only needed to code enough to get a D+ in my Pascal class. It might get great at generating art but I haven't had to generate art since my sophomore year of high school. I hear it's good at scanning lab test results which is great for future me when I need a lab test result but doesn't come up for me in the day-to-day. There are plenty of other useful things I'm sure it can do for other people but the stuff they say it can do for me right now is nothing I'm interested in having done for me.
This is really more a case of possibility #2. Specifically, why are you rolling your own in the first place? The whole reason the output of the AI is "probably flawless", is because this operation already has a million implementations. What's the value in conjuring up another one and adding it to the code that needs to be maintained? Yeah, binary trees aren't terribly complicated, but neither is struct Vector3 { float x, y, z; }; or any of its operations, until you realise you'll be running those operations an awfully lot, so it would be better if it had the right alignment and SIMD instructions were used in its operations, and of course, if you have Vector3 you'll probably also need a Matrix4x3 and a Quaternion etc., so in the end it's just better to use an existing solution.
This is what I mean with turning good devs into mediocre ones: the worst solution becomes the easiest solution, so that's what people will lean towards.
Of course, there can also be situations like you have a very specific use-case and using something off-the-shelf would be sub-optimal, or you're working with old code and making their own was the decision the previous person or your dumbass past-self made, so that's just what you have to work with. But the AI is going to struggle more with those cases too.
Lastly, I also mentioned "energy". If you're trying to put in words things that aren't easily conveyed in English (or any other natural language) all the time, you burn through your cognitive endurance like crazy. This is why you don't want to be in a highly technical meeting in the morning, so being in one all day with a moron you can't really come to common understanding with (beyond the size of its context window) is even worse.
20 years of experience, for what it's worth, and a vehemently disagree. But to extend an olive branch, I think a big reason for the seemingly polar opposite reception of LLM coding online is the context.
If you give an LLM all the tools it needs to write, research and verify it's code, to test the result (for example, in a browser), give it access to all the context it needs, like documentation, use it for code that is new, but not largely novel (CRUD business app vs. invent a new algorithm) and a competent person uses it, it's genuinely great. Great in a way no sane person could deny.
Inversely, if you're working in a shitty environment already, without documentation, where everything exists as implicit knowledge in peoples heads, where you can't test and develop freely, where you're forced to use shit LLMs or "hardened" aka shitty tools, for a extremely uncommon domain, you might have worse experiences.
I completely disagree. I am completely confident that an Opus 5.5 will code circles around the average developer.
I've also had to rewrite AI slop, but I did it with AI. Came out great. Turns out, knowing how to use a tool makes a real difference. It's like saying knifes are shit at cutting because a lot of dumb fucks keep holding it the wrong way.
Yes, the AI is not great at the task without intense supervision and direction. Hence, it's not good at the task.
Example I'm currently working on. Unqualified junior dev yelled at AI over and over to resolve a bug without looking at the code. The AI noticed there was a discrepency between "user_name", "username" and "host_name" fields in the database. The discrepancy applied to two users, resulting from a poorly thought out decision to use "user_name" for a new version of this app. Instead of recommending correcting the two records, it created an additional backwards compatibility layer to pull the entire schema of a table, itemize every possible combination of columns, and write a generic lookup system for every possible combination of columns. This was not even a production system that needed backwards compatibility, it was still in development (6 months of development by this guy, I'm his replacement and finished it in about 6 weeks). And it was not a generic system to deal with that table in general, this was only inside the context of a single auth API method. Without basic intuition, it completely failed to diagnose the problems of its own creation, and just kept endlessly expanding the system to account for its own mistakes that it forgot hours later. The codebase that resulted from this process is absolutely disgusting.
There's a reason for this. It's not designed to do symbolic reasoning. It's designed to passively infer symbolic reasoning from linguistic analysis. It doesn't know what the hell it's doing.
Human slop code existed way before we had auto-complete in our IDEs, way before Copilot start writing small functions and thus bad code is not because of LLM. AI slop is no different than Human slop, it uses a shit ton of ressource to generate it and is faster at getting a working output but bad codebase is not new.
My first couple of programming classes were in Pascal. Then there was the third class where the professors had written their own pseudo language that existed only on the chalkboard and in their own heads. At least I learned recursion from them. I had to teach myself every language I ever used on the job. I was EE instead of CS, so I suppose I shouldn't complain too hard, perhaps there were more practical classes deeper into the CS tree.
Multiple Pascal classes for EE thats interesting. You're not much older than I am and when we did Pascal my teacher said something like its not really used as much now but its a great learning language, but the 2nd class just moved onto C.
I had a friend doing EE and they had Perl as their programming class a few years later.
If I had to guess id say its Python now? That'd be practical as well.
I also leaned every language i used outside of school on my own.
It works great as a fancy search engine. Google results have been atrocious for quite some time, and DDG isn't that much better. When you're looking for something hard to find, you may have to check dozens of results or try several searches. LLMs can do this and process the results much faster than you can, and can even give you the sources. Which you do need to check, because sometimes they still spout garbage.
Most other uses are crap unless it's used by an expert in the subject matter. But search can be useful for anyone. Note: this is not the AI results at the top of the search results; those are still crap. I'm talking about telling the LLM to do a thorough search with a lot more relevant context than the search usually has.
They still.kind of suck gibe you an answer and you say but blah. And itnfoes oh yes in version x.x they changed. I mean if unwanted old outdated in found use Google. Since looking for technical studfnis hard cause so much is out of date.
Genuine question. I'm not a developer, I didn't study computer science, and I'm a pretty basic Fedora user. But why does everyone harp on Linus Torvald's opinions like they're handed down from heaven? Windows still dominates the market, but no one cares what Bill Gates says. I love A Song of Ice and Fire but I wouldn't take writing tips from George R. R. Martin. Richard Stallman's opinions on GNU/Linux are memed endlessly. Linus is literally just a dude.
You should add laptops to the top of your list...cause that's what dominates on productivity...and that's nearly ALL windows! What I think truly dominates is an offshoot of *nix...Android and IOS. Don't you just find it funny when you walk in a mall and see a jacked up Kiosk with a windows 10 taskbar sideways? ...Press any key to reboot!
Asahi is cool but not ready for primetime dev work. Performance is ok, hardware support is spotty, battery life is appreciably worse. Any dev with a macbook is running MacOS 9.9/10 times
As someone who did study computer science and has worked in embedded processor space (so one of the people who's written assembly by hand he's talking about).
I've never understood it either. People take his word as law blindly and it's never made sense to me. Yes he has skill, but so do hundreds of others. Yes he's opinionated, that's certainly not a reason to hero worship him.
For ten+ years people have been pointing out he's basically a single point of failure for Linux and begging for a succession plan and he's been too stuck up his own ass to accept that he isn't the only one who cares about all this.
This has lead to some extreme cult following behind him for some reason.
I've translated assembler by hand and punched the hex codes into an EPROM programmer's keyboard. This would appear to make me qualified to say that he's not always right and, in most cases, fuck AI.
As far as I know no. The plan is for the the top maintainers to gather and choose a successor. Though if you have more up to date information I'd love to see it. I could be behind the times on that.
As someone who did study computer science and has worked in embedded processor space
Using an argument from authority to denounce someone in authority and their arguments? Cool. Might've worked if I hadn't seen you mention recently that we shouldn't use AI 'because Trump and his cronies do'.
Because he's not some guy on the internet saying his opinion, he is maintaining the most relevant Free Software kernel. So if he goes nuts, there is great potential for great disaster. I don't think there are Linux distributions that use forks of the Linux kernel. (What notable forks are there anyway?) At the moment, they all depend on the original Linux kernel. There is currently no alternative. It would take a lot of disruption and a tremendous effort to establish and switch to a fork of the Linux kernel that is free of Linus' influence. Today, I'd say jokingly I will switch to FreeDOS. But if this sloppification goes on, I guess I will have to.
The difference is that Linus does not actively seem to be part of the cult itself rather than using it as the effective tool that it is for someone who already has his type of skill set. He can actually understand and double check inputs and outputs because he already built a foundation of great skill.
Most of the AI techbro community just wants an excuse to stop paying people anything for their labor and get eternal digital slaves to replace them. They aren't wanting to use it as a tool to help them unravel complexity, they are just using it as a smash and grab for power and control. Linus... isn't that, and maybe it's not great that he's embracing it now, during the height of controversy. I kind of wish he had waited for the crash an mania to die down a little and then show it as a useful tool for people who already have skills.
LLMs are useless for a lot of shit but they are pretty damn useful for code. I think some of his recent work finding bugs in old code is evidence of that.
I don't follow him at all, but I'd guess he's making a statement about it now precisely because of the mania and people constantly asking him about it.
Because he still the main maintainer and a developer on a piece of software core to a lot of modern computing.
If he is wrong on this it could ruin the credibility and reliability of Linux and cause a fork that could tear the community and ecosystem apart.
If he's right he's proved that AI can make maintainable, reliable core infrastructure code and the field will change to reflect that. Can't tell your boss AI sucks and is unreliable and I won't use it if AI code is working well at the base of modern computer software.
Because then I'd never finish my masterpiece. /hj But more seriously, a lot of his stylistic choices are ones I try to not replicate in my own writing. The extremely detailed descriptions of food, clothing, woman's bodies, etc. I love the story he tells but find the writing itself can be a chore to read sometimes, if that makes sense.
Most? Where's the data? This hard-line anti-LLM guy has a degree in computer science and has developed neural network systems for a previous company. Many on here that I've talked to about it fall under similar category.
Proving yet again AI bros have lost the ability to think.
"many I've talked to" <- that literally is the data. I'm the data. It doesn't matter what the number is.
As opposed to your comment
none of the most hardline anti-AI folks on here are
Literally me existing already disproves your comment.
here is the email this screenshot is an excerpt of; the specific "X" (the thing which the message Linus is replying to here suggested perhaps AI should not be used for) was the writing of commit messages.
I think if I had taken an icecream scoop to my forebrain I would have experienced more intellectual development than reading the contributions of the sloppers.
You all need to read about enclosure and learn to make arguments without analogy.
My understanding is that AI/LLMs can generate code that works, but often that code is more bloated/convoluted than code written by a decent coder, making it less efficient and more difficult to maintain. I can see that being acceptable for many in situations where hardware/performance/maintainability aren’t a concern like a one-off app or proof of concept. But for something as low level as a kernel I would think that performance and maintainability are the most important things since slowdown in that part of the system ripples up to affect everything a user does anywhere in the operating system.
I understand that a skilled coder can take AI output and clean it up into something at least approaching the quality of their own unassisted code, and I’ve been led to believe that Linus is able to do that for himself. But I fear that other/future maintainers of Linux won’t match that caliber, and it will start to feel that AI bloat. I think I’ve also read of Linus complaining about a torrent of low-quality, AI-based submissions and so I also fear that this will lead to him restricting who can participate in the work, overcompensate, and end up in a situation where Linux becomes less open and more closed.
So here's how I see it: There is a big difference between "AI generated" code and "AI assisted" code. Or just anything.
To put it in terms most people can more understand let's say you were writing a book. If you have an LLM generate the book its gonna suck and be slop. But if you write the rough draft, then paste it into an LLM and say, "Look through this and find typos and other mistakes and point them out to me". Then you go and check the things it finds and fix them yourself, that is fine. It's just a really good spellcheck at that point. Even if you say "suggest a fix" it's fine. For coding you could even tell it like "Take this line of code and suggest 5 different ways to do the same thing that might work better" then test them all out and see which one you like best.
I have personally used an LLM to help clean up my dotfiles for example. I don't want to read through all that shit and find the old junk I need to remove or change. So I just throw them at an LLM and have it point things out to me. Or say you know you have a file but don't know the exact filename. Go into the folder its in and dump the output of ls into an LLM. Tell it what file you need vaguely. Have it read them and give you the correct filename. Then check it yourself.
The problem comes from people who just generate stuff and don't do the work of double checking it themselves.
My understanding is that AI/LLMs can generate code that works, but often that code is more bloated/convoluted than code written by a decent coder, making it less efficient and more difficult to maintain.
This was more true last year but developments in both the frontier models and the harnesses used to control and direct those models have gotten better to the point where I'd say the code is better then most devs. People on here don't want to admit it but if you did a "coding turing test" between Claude and some random dev coding by hand I'd bet most of the time a dev would say the AI output is better.
People said this a year ago and I tested it then and it wasn’t great. I’ll believe it’s better but it’s hard to believe it’s “fixed”. I’m sure I’ll get a chance to try it again sometime. No way I’m paying for it myself lol.
He doesn't get it... Have the big corpo-shilling libertarians rotted his brain?
Like, imagine if using "AI" consisted of whatever big corpo SaaS you subscribe to sending your prompt off to someone to work on then send the result back to you. Isn't that an idiotic dependency to have? And why the fuck even go through the middleman? Oh they have the "best workers?" Fuck that. Nvidia makes the best cards, but you don't see me buying from them.
Do people like Linus not realise how silly it is to offload your work to some middleman service that just steals work from people who do not consent to sharing their labour? Who are desperately throwing insane, GDP-level boatloads of cash at it this unknown technology in the hopes that they'll somehow get people hooked on their particular chatbot?
The intentions of these companies are not good. Their entire valuation is not based on being a "force multiplier" or whatever bullshit, but on being able to threaten or fire workers if they don't slot themselves into the newest enunciation of hype-based capital.
Anything you do to willingly boost or massage these corporate services is enabling actually vile companies to continue puffing air into their stupid pointless bubble until it bursts and the next generation has to eat tuna helper sans tuna just to be able to afford to live.
What is stupid is referring to AI as if it was just one tool when actually it's a bunch of tools, some of which are easier to use sensibly amd some it's near impossible to. Fancy auto complete like it's being compared to sure, fully automated do what the fuck you want agents, yeah sooner or later that is going to go wrong and no one is going to know where or why.
They might be fun and interesting, that doesn't make it a great idea to go all in in my opinion, certainly not for anything you would consider important.
Do you... think it matters? Tool generates code, you look at it, check it and correct it. You send it to a maintainer by giving your name and email, you are now expected to help maintain this code if it gets merged, it has your name on it.
Explain to me why writing the code by hand or with a "deterministic" alternative why would change the quality of what you do or are expected to do in future.
writing the code is the easy part, having a good mental model of what's going on under the hood and planning ahead is the real challenge. if a chatbot puked out the code it's easy to assume that it's right because it worked under your conditions but you might miss stuff that you'd otherwise acknowledge if you wrote the code.
using the chatbots directly discourages thinking and you're deskilling yourself in the long run, depriving yourself of hands-on experience that you'd usually pick up through your coding career.
having a chatbot generate code for you also usually goes hand-in-hand with writing more code faster, which tends to be bad if you like stable software and want to prevent maintainer fatigue.
and that's before we get into the ethical and legal problems.
He compared an LLM to a compiler suggesting that the concerns about using them are the same as suggesting that everyone should write in assembly.
The comparison is invalid, since one mode is deterministic, while the other is not.
LLM-driven development leads to programming via the abstraction of the model, disconnecting you from the source to the point where you don't understand what it's done or why. He's suggesting that this isn't a problem because an LLM is just like a compiler, which suggests that he's arguing in bad faith or doesn't understand the "tools" he's using.
You don’t need to check that the machine code your compiler made matched the code you wrote. You don’t need to check that the commit message an editor sent matched what you typed.
You absolutely do need to check and correct everything that comes out of AI because it’s at best like 80% correct, and that’s if the subject matter is easy.
He's not calling people stupid for not using it, he's calling people that call for blanket bans on anyone submitting code that was made with AI stupid, since it should be judged off quality instead.
Man I've seen others try to twist those words the same way even though it's on the same level as one of those elementary reasoning problems you'd find on a basic IQ test (Jack caught an std after interacting with koalas - a) all koalas transmit stds, b) some koalas transmit stds, c) pandas are free from stds, d) Jack should spend more than $5 on his hookers).
So unless you actually are stupid and this is why you're choosing to (defensively) misinterpret his words by picking the equivalent of c) here, it's clearly wilful misrepresentation of a very reasonable statement - "Of the people I admire, most don't shy away from cutting edge sutff"
This would be true if we were reading a paper on formal logic and all people communicated exclusively through lean.
Why would he chose to include that section in this correspondence? Just as a non-sequiturs? Your analysis implies he had absolutely no rhetorical goal in including that section. Humans tend to have rhetorical goals when they write or speak so we must ask "what was Linus' goal" in that section.
Given the context, the fact that the email thread is in regards to letting LLM's make commit messages, and Linus being famously willing to insult the intelligence of others it would be foolish not to assume that "insult their intellect" was not a rhetorical goal.
So why did he include that part? No particular reason? Was he just making an unrelated remark about some of his friends? Accidentally included that bit but it was intended for a different email?
The simplest explanation is "similar to all the other times he's made back handed remarks or direct statements criticizing the intellect of who he was speaking to, he was criticizing the intellect of who he was speaking to".
This is an odd hill to die on. Linus is famously callous and rude in the Linux mailing list so to declare that this time he was just making a completely off-handed and unrelated remark about his bois is a difficult position to defend.
My last use of AI is in my post history and was fairly down voted because I was being foolishly ironic.
This said, I don't really disagree with his take here despite keeping away from AI as much as I am reasonably able. I do my own critical thinking and hate that I have to jump through hoops on the internet to have a simple question answered without some layman's AI summary that may or may not be even remotely correct or actionable. Even still, when I asked it to check on local real estate laws and provide summaries of possible actions based on a very strict set of parameters, it gave me information that my wife and I were able to bring to a human agent and have things work out better than we had actually expected.
I think a better way of putting it - and I'm still gonna ruffle feathers - is that AI is becoming the new electricity.
Put down your pitchfork and lemme explain.
When electricity came out, people rather reasonably pointed out that simply coming into contact with an electrical device could result in death if one wasn't careful. On top of that, the infrastructure required for it was expensive, unsightly, and at the time incredibly disruptive to daily life. It was going to ruin and disrupt entire industries that had existed for centuries because it could do things a human could not do unaided.
All of this is to one degree or another entirely true even through to modern times, but electricity has become a ubiquitous resource that humanity would now deeply struggle to live without. I won't dare compare AI that far as i genuinely believe humans would be significantly better off without it, but the initial Luddite response to electricity has a historical echo here.
I think Linus' point is more that it can undoubtedly be a useful tool when used in a measured, controlled, intelligent manner, and that it is a genie that is very clearly unwilling to be put back in its bottle. Like it or not it exists, and while we can absolutely be unhappy about it, being blind to its actual real world beneficial uses because the companies that make it are actual festering waste is a bit silly - especially when in the hands of the right chaotic neutral parties these companies are about to have a massive lesson in just how messy their tools have made things.
The problem with the tools , it's not gonna help amateur or rookies to develop the skill without the tools, it seems make the illusion that it is capable to do so but it's not.
While for the veteran or those who know what they doing, of course it will produce various result , from diminish their entire person or just become another tool they wield for their work, depending on skill issue.
Every once in a while I'll ask the DDG AI a question, and then try to confirm it with narrower searches. I think that's the extent of my use of LLMs.
And I feel guilty for it, wondering how much water and power it's taking and if the compute is poisoning the air of its human neighbors and keeping them up at night with a 70 dB hum.
There might be something to be said about using models locally, and using them as part of a human-guided process.
Note: I've stopped calling it AI in favor of LLM, partly due to the whole SI controversy in the US and partly in the hope that AI is still a larger category of learning systems that might lead to AGI, if we don't destroy ourselves first. Also the Jev model looks like an interesting development that indicates the ultimate tools might be other than a later model from OpenAI or Anthropic.
Note Also: I programmed in the 1990s and broke my brain trying to teach a computer to competently play a complex strategy game. I got pretty far before all my variable names leaked out of my head. These days, I write simple AHK scripts but that's about it. So I'm way out of my league when it comes to sound 2020s era coding methods.
Yes but unfortunately this specific tool is here to stay. Either we learn on how to use it properly and ethically and we regulate it, or we will be in deep troubles. Just denying and refusing to accept AI is here is dumb.
By regulate I mean at state and super-state level (not thinking of USA, but of the world) and define and accept proper ethical rules, both for individuals and corporations.
That blog post is just a bunch of insults to anyone not agreeing with the writer without actually giving any elaborate or truthful argument on why is not just a tool and it's always inherently "evil".
Some arguments made can be easily disproven. For instance the climate change one.
They main point of the post is that "no matter how you use LLMs". But then the argument given here it's highly dependant on how you use them. The amount of environmental damage of LLMs completely depends on which LLMs you use, how you use them and how often you use them.
This just here it's showing the irrationally of the blog author.
But it goes one, next point is economics. Which is also 100% dependant on HOW YOU USE IT. Not all LLM are made by big evil corporations, and there's nothing in the technology preventing it being made by anyone.
For a blog post whose main argument is that saying "it depends how you use it" is wrong, have only give arguments that are dependant on how you use. The author makes broad ignorant assumptions about how people use LLMs, asserts that everyone uses them like that and then pretends to control everyone lives based of that. That way of thinking is highly dangerous and it's what have given birth to a lot of evil on our world. We need to be better, more rational, more open minded. Not falling for easy politics and actually do complex political thinking that won't give simple answers to complex questions.
These research papers talking about AI leading to cognitive decline seem so asinine when certain people's productivity has just skyrocketed under AI.
Sure the hand coding capabilities decline, but that's replaced by thinking around the actually product or architecture.
A huge sect of new Linux users and Selfhosters are talking about how much AI has enabled them yet we're still bringing up papers on cognitive decline that were written on models from 3-4 years ago when Claude Code didn't even exist.
Those papers, and the most recent ones, all point to use of LLM's as causing cognitive decline due to offloading problem solving.
Saying "they don't even take into account this way more capable model that makes me able to solve problems way faster" is like saying research into cancer causing materials is bunk because they don't take into account all the times I inject concentrated carcinogens directly into my testicles.
No one is saying using LLM's slow you down. They're saying that being slow was a feature not a bug and skipping it prevents long term learning. Is there a way to use LLM's that improve learning? Maybe. My worry is that even if there is there is little real inventive to do so.
Yeah even writing lead to cognitive loss, because now people don't have to memorize everything.
It's just fear for the future and for change. And this is probably the most massive change in human history since fire was discovered.
But don't fear too much: the change is here and will stay, that's the only sure thing
Edit: what I mean is that accepting the fact AI is here and won't go away because we don't like it is the path to start regulating it at government level and agreeing on ethical rules to use it. Just saying you are a bad person of you use it it's like hiding under the sand. You lose.
Got something that has more than 18 participants split into 3 groups for a total sample size of 6 people per control group? Not to mention its focus on AI assisted essay writing, measuring cognitive debt (not decline), using EEG to infer performance of higher cognitive function, etc.
No. AI is simply too recent and evolves too quickly to be demonstrably anything.
The studies you refer to are mostly flawed, and most importantly not about software engineering, where I assure you the problem solving part is not diminished, simply shifted. Don't let the memes that show people type "just do your best bro" into claude code influence you, they are just memes :)
Linus Torvalds is not an intelligent person. His argument style is extremely feeble, propped up solely from an ulterior, omnipresent appeal to authority. If he was just another poster, we would handwave his politics and social arguments as utter drivel.
He created some awesome software. Git, Linux. But the dude should have been kicked out of Linux dev a while ago. He is no longer required, and just runs his mouth constantly.
I wouldn't ve surprised if this was generated by Claude.
Why yes, let us compare corporate, proprietary, closed-source tools that inherently rely on environmental destruction, breaking licenses, and intellectual property theft in general, to basic, completely free and open source programming tools that have been in use for 50 years, and harmed nobody during their development.
Of course this guy does all his work on a MacBook. Because of course he does.
I've usually been with him, because he's very knowledgeable, and quite often his takes only seem ridiculous out of context.
But this is borderline corporate propaganda. If it continues, I'm afraid we'll all be moving to BSD.
He really thinks the tools you use do not shape and limit your results and yourself. I'd take a book carefully crafted in LaTeX anytime over slop pasted from Copilot into Word. What he's not even seeing is that AI is, above all, a source of input. No other tool, not the best LaTeX editor, takes care of my writing. AI is totally different. It is the author of Linus' code, it is his ghostwriter. He did not write it. He should not claim authorship. If he's not the author, he can't legally release it as Free software under a GNU license.
My calculator can do math for me, but I need to know how to use the calculator, the operations, rules in math, and basically, well, know math. My calculator is a good tool.
"IA" can "craft" a magic result without you needing to known anything about what the tool is helping you to accomplish. In a perfect world IA can be used to automate the most tedious part of any job (scaffolding, auto generated code for APIs, color grading, laso tool, etc) but in this world people use it as an excuse to not learn what they are trying to do. "IA" is a bad tool. Is not helping you, is doing it for you.
Never in human kind has existed a "tool" that does things for you without the need of any input or knowledge of the tool itself. "Engineer" comes from "who operates the machine", "IA" operates itself. I don't think there exist a real comparison with that "IA" does, but it's effects are clear, and are not good.
30 years ago when I was in school, cheating was rampant. There's going to be about 15-20% of people who understand when they are cheating themselves, and the rest won't and never have.
I can think of plenty of tools that will do something for you even if you understand nothing about the tool, how it works, or the domains it is helfpul with.
I can plug a drill in a socket, put a random bit on and drill a hole without any knowledge of how a drill is supposed to work. But my hole won't be straight, will be too deep ir too shallow, too large or too small, and I might have drilled straight into a pipe. I don't understand how drilling works, I still drilled, but the outcome is shit.
I can mix a bunch of flour, eggs, sugar and chocolate in a bowl, then put it in an oven at max temperature on a random heat method and wait half a day. I'll have a bunch of coal, not a cake. I used an oven without any idra how an oven works, it did something for me. It's shit, but it did something for me.
Anyone can put the key in a car and start pressing random pedals to figure out how it works. Eventually anyone will get it... If you're not dead before then or you didn't total your car. Without understanding anything about a car, I did manage to move it and go somewhere. But I damaged some things in the way.
I could go on and on with as many tools as you want. But I doubt you'll keep reading.
LLMs in software development are the same: If you understand how to steer it, if you understand software engineering, it will have fantastic results. If you don't, it will have shit results. The only difference is that it is much harder to understand the insides of how an LLM model works, as that is another specialty. But just like you don't need to understand how an oven actually works to use it, you don't need to understand how an LLM actually works to use it.
Edit: Another great example is a camera. Anyone can take a picture, but without a good understanding of photography, most people's pictures will be shit. And the vast majority of users do not understand and do not hace the knowledge to understand how a camera works.
I agree. The main issue with AI in that respect is that there's no easy way to know how much it actually did vs how much value the user added while using it.
Things made mostly with AI are terrible in the long (or even short) run, but if important phases like analysis or design are mostly done without AI, the results are great, since it saves a lot of time for specific tasks.
I feel like this is the same as any tool, with the difference being that someone can fully hide that they used AI without it being initially evident, or lazy people use it as a shortcut to avoid learning. As long as you got all the base knowledge right and guide the AI yourself instead of letting it control you, it has potential.
Also, AI slop is so common nowadays that there's a bias towards them and all projects that use AI get labeled as bad by default, which doesn't help either.
You can know a hole is not straight. You can see it. Besides, if your house is built by someone that doesn't know how a drill works, well, it's going to crash down over you. Then why are we accepting works of people that can't and doesn't know how to properly use the tool?
Now, both of your examples have a person having to use the tool and know how to use it... IA is having the drill doing a hole by itself and just say "good enough". The house will crash down.
My point stand against both your examples. Both requires a human doing the work and the human to know how to use it to be useful. If the difference between a surgeon and a butcher is precision and knowledge, why are we allowing the butchers in on everything?
"IA" at best is a tool and at worst is a substitute. A "tool" that doesn't ask anything from you is not a tool.
I stand by it's not a tool. It's a substitute. But even by labeling as a tool, why should accept works from people that doesn't know how to use it and when? By following your examples we can't trust anyone claiming "I did this" with an "IA" until we or a professional inspect it, and since you can't inspect the absurd volume of things "IA" spits, the only solution left is ignore it all. Either you trust blindly on the guy that says he knows how to use the drill and risk your house falling down, or only accept works of a professional, and if you only take the job of professionals then what's the use for "IA" then?
Slop harder, Torvalds. You fucking lost it, and you're ruining Linux for everyone.
At least the other kernel maintainers still check things and catch problems, because clearly Linus Torvalds doesn't care anymore and would rather make false equivalence excuses.
I've used Ai to try to make me operating systems and programs. All it does is give shells that you have to code the rest of what you want into it.
Then again, I don't pay for Ai and I've pretty much exhausted everything I wanted to ask chatgpt within the first 4 months of using it. Then I was able to make gemini think it was a different universe for a while until it caught on. Then I stopped using it as much.
It was fun for a while, Ai is still something you have to pick and choose and use good judgment with. It's not an end-all of coding.
Can't help but laugh at you guys correcting literally the undisputed #1 software maintainer in the world. AI is just a tool - if you fail then it's a skill issue.
He used to be cool, but then he started following some corporate hype trends that make me suspicious, I think he is nearing retirement and doesn't want to fight corpos anymore.
The moment he accepted Rust officially on the kernel was the time I knew he got a bit too old to care anymore.. This is just a continuation.
Also is so curious to see the venn diagram of people supporting Rust and AI being mostly a circle. My theory is that Rust takes every piece of joy out of programming and thus the idea of never programming again becomes so attractive if you are surrounded by Rust.
The Luddites weren’t opposed to technology. They were opposed to, in their words, “machinery harmful to commonality”.
It’s a shame that their name and cause has been totally contorted by the capital forces that opposed them, cuz it’s actually a pretty good lens for analyzing the societal impacts of AI.
Very interesting. Great knowledge, thank you. I guess I’ll have to revert to calling it being like the Amish.
Linus is talking about the worker using the technology. He likens it to other tools used in his trade. I don’t see him arguing about who should see the profits from the technology.
I think being dogmatic about strictly not using AI is not empowering the workers who would benefit from utilizing the tool in the correct context.
Luddite is correct. It just isn't the insult you thought it was. I'll add to the above commenter by saying that the Luddites were textile workers, which was the most highly skilled labor force in the world up to that point. They had I think a 6 year apprenticeship to learn to work with their very complex machines. The Luddites were the software engineers of their time.
They didn't want to do away with the machines, they wanted the workers to be in charge of what machines were used, when, where, and how. They were specifically concerned with the markedly lower quality of the fabric being produced with the new machines. In the burgeoning industrial capitalist economy, this could be considered an early example of enshitification. The Luddites were a labor movement concerned with the bosses ruining their craft with cost cutting measures, and they got branded as strict anti-technology primitivists so hard that it's still the common cultural understanding today.
There are of course no modern parallels for this and anyone skeptical of AI's usage today is definitely a primitivist that wants the AI genie put back in the bottle. /s
To quote Cory Doctorow "What's important isn't what a machine does, it's who it does it for and who it does it to."
Yeah that’s a great reply if you were trying to competitively comment but 99.9% of the online discourse against ai is completely against it. Including most of the discourse on this post.
Luddite is not the correct term. These folks are more dogmatic against Ai while Linus is literally talking about workers using it as a tool to make their work easier.
The work Linus is referring to is authoring commit messages in the Linux kernel.
He's arguing for letting the AI write the commit message. Honest question for you; if a person does not have the ability to articulate something like a change to the kernel, how do we know they understand the change?
He's a hero to a cult of tech bros. He's never been my hero and there's many in the Linux space who have been calling him out and asking for a democratic succession plan for ages. He's not some all knowing perfect being.
Sort of. The plan is "convene to make a plan" 72 hours after Linus's git stops updating. I should have been a little more clear though that people have been calling for a democratic secession plan to happen now instead of when Linus passes or steps away.
So not being thrilled about huge corpos throwing billions of dollars destroying the environment, ruining people's minds and jobs and potentially killing us all in 10 years is a "dogmatic Luddite stance". Got it.
That’s weird, I didn’t see anyone approving of those things in the original posted text and I definitely didn’t imply any of that with my comment.
Unless I was mistaken this was about Linus saying that using ai is akin to using any other tool.
Actually with some thought, you’ve convinced me.
Hell the printing press is being used by our oppressors to kill the earth. I will be deleting my account and from here on out I will only communicate and read through hand written notes.
Amish is also wrong. The Amish aren't against technology which is why you'll see Amish communities using things like credit card readers, industrial motors, and even computers.
The Ordnung, an Amish community's governing body, will review proposals for technological adoption every 6 months and hear arguments for and against adoption. The decision to adopt is screened against potential harm to the community based on a number of factors.
Honestly, the model the Amish use would be an excellent way to achieve a pro-social form of what the Luddites wanted to achieve.
"Y-y-yeah well. If you like em' so much why don't you marry em'! Hehe got em'"
Please let's try and keep this above middle school. The whole point here is that comparing people who fundamentally disagree with this tech as "Luddites" or "Amish" betrays a lack of understanding of those two groups of people.
The Amish are modern day fools who reject this very thing we are on. I still say if you are so ardent to defend them you should abide by their rules and log off.
Your fake quote was a zing straight to my core. I’m not sure I will ever recover.
I didn't know Luddites were an actual thing back in the day, so i guess i learned something today, but according to wikipedia these days it's used as a term for people who are against new technologies in general, which isn't accurate cause i'm not against new technologies in general, i just have an issue with genAI in particular. I don't see how Amish would apply here either. Amish don't have a problem with specifically genAI, they just reject most things from modern society, including a lot of technology.
505 replies
Yeah, but your compiler wasn't made with massive copyright infringement, stealing much needed water and rampant corruption, was it?
I guess Linus doesn't care about the environment or any of the other issues surrounding A.I. I'm not anti-AI just for the sake of being anti-AI. I have a huge problem with AI bcz of it's environmental impacts, intellectual property violations, job impacts, attempts to replace all employees with AI, the general dumbing down of society and worse off the internet is bcz of AI, the surveillance state being built around it. I could go on, but you get the point. I am a staunch, no AI person bcz of all those reasons and a hundred more.
This isn't an age thing, or a, I am getting older and Im stubborn thing. AI is literally ruining the planet and everything on it, and this guy thinks we should just accept that.
I say fuck off with that noise, I don't have e to accept anything.
It's perfectly ok to use a new tool to do your work faster, more efficiently, and with greater precision if these improvements are justified by the cost of the tool. But if that tool is a coal-fired, water hoarding, ophan crushing, billionaire-owned kleptobot, your results need to be better than pure gold to be worth the true cost of using it.
I don't care about Linus, Stallman's opinion is the interesting one.
Those examples are all a set of mechanical rules that will generate the same output every time.
You don't even use auto completion without checking what it actually wrote, right?
Linus is a guru engineer. His judgment of things outside of that scope may not be quite as sage.
Yes the tech can be useful. But should it be used? The former is a descriptive statement, the latter prescriptive. They are not the same.
I've thought since the beginning of all of this that the biggest mistake was calling all of these LLMs and other tools that they smashed together "AI". It's misleading and ruins the reputation of all of the tools. If someone had made a coding tool that ONLY did what "AI" does with coding, no one would give a shit. But now because an LLM spelled "Strawberry" with too many "r's" well that means it's all slop and it's all shit.
As for the people throwing shade for ethics, EVERYONE breaks their ethics for consumption and EVERYONE has their justifications. I'm not saying that environmental issues aren't important, they very much are, but giving others shit for their choices when half of your clothing and your phone partly relies on slave labor and stripping of the environment weakens your attempt. It's easy to sit on a high horse and judge others for their use of AI. It's much more difficult to understand the nuances of why they do. Just like you would want someone else to try to understand the nuances of why your shoes cost $150 and the 9 year old who made them was only paid 2 cents.
The point that I'm trying to make here, is that you can complain about AI without shitting on others.
Linus become like one of the boomers excited by some dumb ai slop on Facebook
But worse he doesn't care about how unethical is the present ai
I lost respect for this guy
"the kind of expectations i have is that people do their best and i expect people to understand what they are doing."
my immediate response to some idiot thinking i accuse him of cheating because i voiced concerns over his methodologies: "look, if you want to drink mercury while writing code and dont see any problems with that, fine. go ahead. waste away your health. but at least have the decency of sitting under the venting hood and stop making it everyone elses problem."
When the topic is AI, the quality of some of these Lemmy threads drops down to the level of YouTube comments.
I got banned from the Fuck AI community for saying Linus isn’t stupid. Not that he’s right on AI, just that he’s not stupid.
Indeed, the atmosphere is toxic.
We’re on a site where people happily discuss their arr stack or transpiring the PS5 system, but bemoan AI is killing artists and coding.
There are real ethical issues, job impacts, environmental issues, property rights issues (real and intellectual), privacy, a stack of cards economy etc.
But there is also utility, and frankly the genie is out of the bottle. For better or worse.
Catching kernel bugs in Linux with AI should be seen as a good because they’re not being exploited offensively by private actors. AI review that catches the most annoying 80% of errors or issues saves human time, while not contributing slop to the kernel.
There is a lot to discuss if you show up in good faith.
Also, Linus is an asshole. Why would anyone be surprised he reacted like an asshole? His point on his expectations is actually clear and sensible to me. He probably could have said just that, but he’s an asshole, and acting how you would expect after a bunch of people are yelling at him.
Nuance is a hard concept to get in most Lemmy comment threads. It's obvious it's just a tool that uses an insane amount of resources that can do some good in certain applications. The problem stems from it being abused for evil and stupid things.
Can AI be used to help flag cancerous growths in x rays and help save lives? Absolutely! Can it also be used to track every human being on the planet for surveillance purposes while its proponents build massive biome-destroying data centers to act as the infrastructure? It sure can!
All that said, I still don't feel confident about rolling in AI code to open source projects when bots aren't accountable to their contributions. They should require a huge amount of oversight which I doubt is happening at the moment.
I want to expand on your first thought and just say that in my experience the internet as a whole is not a great place for nuanced takes, and I've noticed that becoming mirrored in reality over the last decade. Perhaps its been happening for longer.
There's so much going on, and so many actors trying to influence the masses. I don't blame anyone for having different conclusions to my own, there's an almost untrackable amount of sources to draw information (and disinformation) from. I just hope we can get to a point together where we can (again?) discuss nuance without ceding to emotion.
Agreed.
On AI contributed code I’m always confused, if I put MY name on it and push it I am stating it meets my standards. If I allow a Claude contributed commit into my code, I am stating Claude passed my standards.
If someone is contributing and it’s not up to standards, why is it getting merged? If it’s a maintainer and they’re merging crap with review or standards then you couldn’t have actually trusted them pre-LLMs.
Ever since the first iteration of GitHub Copilot it was always your responsibility to make sure contributions to your project were correct. The beauty of open source is that when a project jumps off a cliff you can fork it or make your own.
Partially what is good about Linus is historically Linux has said no to crap, and the Linux team expend considerable effort to remove crap they realized was crap too late.
From my layperson understanding, the true program is not the lines of code, it's the model that lives in the programmer's brain. The code is just an extension of that.
If it needs debugging, it is the mental model that helps. If you need to add a feature, same thing. If a few years go by and no one remembers how the code works, rebuilding that mental model is the first step in making progress.
I suspect that it used to be that if someone contributed good code to the kernel project, it meant that they had a sufficient mental model. But with AI, it can seem like they do.
Maybe if you're Linus your understanding of the kernel is so strong it would take years to atrophy, if at all. I don't know. But developers in general?
Too many times in my work life, people have told me one thing, but it turns out that AI invented that and they didn't really know what they were talking about. On things that I'm an expert in, I can catch it. But if I'm not an expert, I don't realize the mistake until a consequence occurs.
Allowing AI is basically making it so the maintainers have to be complete experts in everything they accept. Maybe that was already the case, but that is my layperson understanding of it.
This is correct. And yes, we expect everyone that contributes to have this understanding. If you don't have it, welcome to ask. Misunderstandings that happen despite your effort will be caught by the steward.
This comment disguising itself as nuanced by saying the word nuanced a bunch is really nuanced.
Just don't engage with any community that starts with "fuck" if you have any kind of nuance for the subject matter.
It's not a great time. If you don't get banned, you'll get dogpiled with angry and idiotic comments anyway, by people that have decided their own facts already.
All the fuckX communities are toxic for the same reason.
No nuance. No complexity. Binary.
The world is not black & white it's shades of grey.
I got banned from saying not all that is called ai is bad by giving video game npcs from the last decades as an example.
I thought it was just llms that they where against but nope. I imagine Alan Turing spinning in his grave.
Vague criticisms, fallacious reasoning, highly emotive language, little response to attempts at discussion (though some might engage in disingenuous discussions of shifting goalposts here and there to discourage particularly vocal engagees from future participation). So I'd say that Linus is just the latest convenient scapegoat in an Anti-AI 'trend' that will have little impact on the rate of AI adoption in the long run.
Sucks that I waited until now to try to form a measured opinion on AI because now all the discussions on the internet are hot garbage.
Discussions on the internet are hot garbage in part because thanks to LLMs you can't know what's real anymore. Of course there were bots before, but any random idiot can pay OpenAI for a subscription now and spam anywhere they want.
Thanks for your contribution to the pile.
Measured perspectives are hard to come by. A Marxist Perspective On AI
There's nothing Marxist in that article. Only fallacies from rich white man ranting about people pushing back against fascism.
People say AI to mean LLM first because main stream western, fascist, media pushed the term by inviting exclusively non-expert and CEOs to talk about the subject. People are pushing back against a product that is mean to make workers dull and replace them by
stealing their works even by capitalist standards-> exploiting their works, appropriating it to themselves without any compensation nor permission.The article tried to paint the discussion as anti-llm would be anti technology, which is not Marxist in any way and not relevant. Weirdly enough fascist guy like the one on this article don't talk about cryptocurrency as something we should chill for. Isn't it a tool too?
Current LLM chill is the opposite of dialectical materialism: what is the concrete analysis here? LLM outputed nothing of value for years despite the hundred of billions invested. Worst it actually participate to global warming, bourgeoisie attempt to mass control, getting rid of workers and inflating computer costs. So the real consequences for people are losses only.
And for what? To get to have unskilled rich white people who can afford the awfully hight price of LLM to boats about "I made this" while it is actually made from steal and empoverish people and enriching tech lord's.
So no, thanks, and trying to paint fallacies to Marxism isn't going to help either.
EDIT: replace the "stealing" argument as pointed out -> https://lemmy.ml/post/54102413/28384201 , this can be confusing with the anti-piracy propaganda vocabulary.
Stealing: when you don't have it anymore afterwards.
Copying: when both have it afterwards
The RIAA's propaganda campaign was sadly very successful.
You're right, I said "steal" as it was the most common word, the right one was "exploiting". Fixing it.
Actually let me analyze again your comment here: I thought you were right at first but thinking about it not really.
First let me address : I didn't know about RIAA, this kind of paternalistic way of addressing people is not good in any way. We aren't all USA-centric (like you?).
Then about the "steal" argument:
Downloading this comment is not authorized.
You wouldn't download an argument, would you? Or would you? What, you're reading this in your browser without authorization from me?? FBI, open up!!!
are you trolling?
I checked your history and you seems to genuinely interact but you're sentence makes no real argument (more like a joke).
But maybe you haven't read the rest of the conversation so here some points:
Private property is theft. "Intellectual property" is a form of private property. It exclusively benefits the bourgeois.
Regarding your third bullet point: Try your example again, but this time the boss doesn't copy the file. It's still labor theft. The copying is irrelevant.
Are you saying I am right then? I feel you think of something but it is not clearly getting to me here.
it's labor theft-> hence the termstealI was usingcopying is irrelevant-> so I was right? the labor theft of big tech in order to sell LLM is wrong on many level and we should fight against it.Do you at least see that, to me, what you just said is validating my points?
Also if anyone else (not to you @m532): "AI" (in reality chatbot or generative AI) should be made by the people for the people, and FOSS ideology (can we say Librism in English?) have always, and still mostly is, against the closed-source (close dataset to be precise here) nature of current big-tech AI. This was already a problem 8 years ago when the space started to get filled with "scientific" paper that didn't disclose code/data (sometime just code but still insufficient) hence couldn't be reproduce while sometimes having wild claims. We shouldn't give so much credit, attention, money to those fascists in charge of selling as much LLM as possible, whatever human, economic and environment costs.
Bad news, @[email protected], you’re not a Marxist, and your analysis is the opposite of dialectical materialism.
was that supposed to be an argument against my explanation?
Obviously it wasn’t, or I would have made an argument.
Yes please do not make any argument and post memes with some interesting comments
Really helps.
I'll never recover from this. 🤣
We have our share of radicals
I would like to reiterate that posting on main as an admin is obnoxious since we can't block you.
Consider using an alt to post unless performing official instance actions.
If you can’t block me, that’s an issue with your Lemmy client, because the default web client isn’t stopping you.
You're wrong: https://github.com/LemmyNet/lemmy/issues/6706
Oh, thanks, duly noted.
Button is right there man
https://github.com/LemmyNet/lemmy/issues/6706
the button is there, but clicking it does not do anything, can't do it via the profile or searchbar either
What are you using? Web or some app?
just the basic lemmy.ml website.
i personally don't care about posting opinions on an admin account, but they seem to be right. suspect it's to do with living on the same instance.
Cory Doctorow's "Reverse Centaurs Guide to Life After AI" has a short section that pertains to this. He talks about how someone like Linus essentially doesn't have a boss and has untold resources for whatever "AI" "tools" he finds useful.
He isn't ever forced to use the wrong tool for the job or use a tool that simply doesn't do what his boss thinks it does. He doesn't have to worry about getting fired or his coworkers getting fired because his boss was sold a story about reducing labor costs with this cool new "tool". He simply uses what's useful to him with the probably millions of dollars of compute made available to him by the tech giants.
95% of people don't have that luxury. They have a boss itching to get rid of them or 10x their workload with the help of the "tool" that Linus is so happily and defiantly praising on behalf of the monopolists.
I don't see how this means he's wrong though, if people are hating AI as a proxy for hating their abusive relationship with their employer, that doesn't seem like a strong argument that everyone who isn't in that situation must also hate AI.
It doesn't mean he's wrong! it means he's out of touch and doing the propaganda work of capital for free much like many other self-employed/business owning/petit-borgeois people who think they're on the side of capital.
It also means that his praise for and indeed positive experience with this "tool" is largely irrelevant to the broader social/economic discourse surrounding so called AI.
Blind hatred for AI as a technology is inherantly reactionary, but that doesn't mean that the implementation of the technology today isn't problematic in many ways.
Yeah I notice he doesn't address the costs, economic and environmental, of LLM proliferation. Way to ignore a large part of the argument while calling anyone opposed to your position stupid.
I hope he gets to eat those words one day, but I doubt he would ever admit he was wrong.
@[email protected]
I hope you were shamed by this reply, on the use of commas in a subordinate clause.
Hahaha my writing style is runon and compound complex sentences without commas.
Oh my god a reasonable take for once. Thank you for putting into words why his heel-turn has bothered me so much
It's funny, the c/fuckai mod that banned me a while back said the same thing. Iirc it was "How very reasonable of you. Banned"
Then the picture would be significantly changed if the cost of running useful LLMs goes down to what it costs to acquire a reasonable computer to work on the Linux kernel, no? That is if we could run useful LLMs for kernel programming on our kernel programming computers, and we own those LLMs. In such a world, capital would have no stronger leverage over us than it already does by selling us the computers we need to program. Does this argument make sense?
One should look at China when they're concerned about the USA approach.
Yeah, but tbf all of the statements he's making about it are in the context of organizing linux development and the most effective way to carry that out, and just happen to be public and so end up being widely republished and incorporated into discussions with a broader topic than that.
Yeah, I don't think Linus should be policing his speech at this level or for this cause. Educate the workers please, Linus is doing plenty.
Shush! They are not supposed to know that the AI–Anti-Ai fight is fabricated bullshit with both sides being equally stupid (like people arguing based on ideological zeal always are...). And that it's just a show so that the majority is occupied with watching the mud fight so they don't think about the actual problem.
What's the point of this argument though? Sure it's real for every profession and every tool but what's the agentic choice we ought to pursue here? Whine against AI or actually lead society with policy? As "ban llm" is not a valid policy that would ever work so the only viable path is regulation and growing of culture where AI is used for good, no?
Embrace AI, work towards open weight/free software and protect worker rights. Society wins at the bleeding edge not at conservstism.
The answer is certainly not to embrace "AI" in it's current forms or allow for big tech monopolists to continue their wasteful and destructive infrastructure buildout. The oligarchy isn't going to regulate itself to worker self-determination either. Nothing short of the death of capitalism will do.
Ok, let me know how that goes. While you whine online and have no impact on reality we will embrace "AI" and drain billionaire funds for distillation, open weights and building free software.
You genuinely don't see the effective differences of thse two strategies? Actually doing something vs having your finger crossed some magical boogeyman called "capitalism" will collapse?
So he lacks empathy is what you are saying.
I think there's a broader question here from an evolutionary perspective. Humans are fundamentally tool users, and technology has been central to our development as a species. It's possible that this evolutionary niche we've occupied will eventually reach its limits and contribute to our own downfall, but I'm not convinced AI is uniquely responsible. It's part of a much longer trajectory of technological progress.
We've evolved to solve problems and use tools to make our lives easier. That doesn't mean technological progress is inevitable, but these tendencies, combined with competition, make it difficult to imagine us ever really stopping. Even if we stopped today, who's to say another generation of humans 10,000 years from now wouldn't continue down the same path?
I don't think this is exclusively about capitalism. Capitalism certainly incentivizes growth and reducing labor costs, but competition, greed, and the pursuit of technological advantages have existed under other economic systems too. That doesn't diminish the legitimate concerns about how AI is being used against workers.
I also wonder whether our short lifespans make us particularly bad at considering the long-term consequences of our innovations. Who is going to sacrifice the opportunities of their own generation to prevent a possible catastrophe centuries from now, especially when there's no guarantee it would stop technological progress?
I have a degree in psychology and particularly enjoyed evolutionary psychology, so I think it's worth asking whether the deeper problem is our own nature, and whether our capacity for collective restraint can keep pace with our ability to innovate.
I know this raises more questions than it answers, but we've faced civilization-changing technologies before, from the printing press and electricity to nuclear weapons. AI won't be the last. Perhaps the more important question is how we learn to manage these transformations when our tendency to create them seems so deeply rooted in who we are.
This argument is conflating the bad boss with the misattribution of the tool. You could do that for any tool throughout history. But I still fail to see how you go from "bad boss is bad because they make us misattribute tool" to "good boss is bad because he's not wallowing with me about how bad boss is bad".
As anyone who has known about him for more than a couple of years knows, Linus has a pretty solid track record of calling out stupidity. It's just virtually always stupidity that affects the kernel. If he had to stop and tell bosses at other companies how to operate, Linux never would have made it out of the 90s.
I guess I have that luxury, but I sacrificed hard to get where I’m at.
And, truthfully, way more people could do it if they didn’t feel entitled to the easy way (college > world destroying corp).
Entirely irrelevant to the conversation at hand but okay
I like Linus, but comparing an LLM to a compiler is fucking stupid.
Why?
You can predict with certainty what a compilers output would be. Linus is conflating the fact that most people don't understand how a compiler converts higher level programming language into machine code with the reality that no one can actually predict and audit what an LLM puts out.
He is saying that you have to understand your contribution, so whatever the LLM spits out, you need to thoroughly review and double-check that it makes sense. So, in particular, he's not trying to rate these tools, he's merely saying make sure the contribution is objectively good, then he doesn't care how you got there.
Predicting what a compiler outputs is quiet hard, especially if you enable optimization flags or get near any undefined behavior. But I think you're right to say that a compiler is more predictable than any LLM. I don't get the point why you should be unable to audit the output of an LLM? If its used as a tool like described the user should have the knowlage to audit any output
Strangely idgaf about whether he respects me
I don't think he thinks about you at all
👍🏻 The vast majority of people don't.
The other practical side of this is enforceability.
Lets say Linux banned all use of VSCode in contributions, this second. And any Microsoft tools. No exceptions. And it makes sense, because there is an evil blatantly anti-linux entity behind it... What are people gonna do?
Keep using it anyway, and just not disclose it.
This is why linux doesn't tell people what tools they can and can't use. Practically, they can't police what tools people use, just what they contribute. Just like (to a reasonable degree) they don't try to police people who contributes to the kernel, but instead judge them by their contributions. Other contributors don't have to like those tools or contributors, but that is how a project like this must operate.
Much of open-source development relies on the honor system. There is likewise no way to tell, just by examining code, that it was not plagiarized.
Honor that needs to be enforced is not honor. It is something less.
It relies on legality, not honor. E.G. with GPL, or code with a proprietary license. It's the threat of losing their job, getting sued or whatever that discourages plagarism.
Non-point elements removed to help you focus.
Ah I get your point now. Yes, ultimately both policies would rely on honor.
I'd argue plagiarism is different, though. If a codebase stipulates 'don't plagiarize,' that's... common morality. Not many contributors are going to argue 'well, I want to plagiarize.' They wouldn't advertise it in any project they contribute to.
Dictating what tooling to use is different. That's going to fall under 'personal choice' for more contributors. And they may feel compelled to hide tool use for that particular project, when they would otherwise show its use publicly in others.
In other words, forbidding tools like, projects runnings be telling contributors to forgo something they're used to using public. That's going to trigger more lack-of-disclosure and rule-breaking from devs that otherwise wouldn't do that.
hes comparing knowing how open concepts work vs using proprietary tech whose purpose is to hook you and then enshittify after you become reliant on it
It’s a pretty reductive take, that treats tools as politically and behaviorally inert. The work on understanding how cultural artifacts use us just as much as (or maybe more than) we use them goes back several centuries.
But Linus is just an engineer, after all. There’s no reason to treat him as some high-minded philosopher (and he’d probably be the first to tell you that). And for as long as the history of thought into the politics of artifacts may be, the history of engineers focusing only on solving the problem in front of them and not thinking about “where the rockets come down” is about as long.
I'm kind of gobsmacked by this. I've heard younger developers compare LLM with compiler and chalked it up to ignorance, but I don't get what this is.
Can I write code in assembly? Yes. Of course. It's true that I never do it, but if there was a rule that you needed to write some assembly now and then to certify as a kernel developer I don't see how it would be any kind of impediment.
Comparing more directly: if you hand me a compiled binary and the source code and ask "why is the bit at address 0xXXXXXXXXX a 0?" I can, with 100% certainty, answer that. Probably in less than a day of effort. If I ask "why did the LLM output "for" as the XXXXth word of its utterance, can Linus say why?
And "can you write a good commit message without an editor?" Uhhh? At least when I was in school, this was something people did for fun/messing around. I still do
cat > somefilewhen I need to put some quick text together. In fact, Claude will sometimes write a Git commit by cat-ing into a file--which it's doing because there's so many online examples of this. Fucking weird things for Linus to pick. Makes me think his brain's rotted.I'm finding the allure of AI is particularly strong for people who have risen to positions of power where they had to stop doing the work themselves. Instead, they had to work through other people to get bigger projects done.
The allure is that you can just say what you want done, and it does it. If it's done it "wrong" you can just tell it it's wrong and not have to worry about arguing the technical differences. As such you feel like you get closer to what you wanted, rather than somebody else's take on what you wanted.
The things is, when you are removed from the code, you can only work at a high level. That person you asked to do something is considering 10 other low level concerns that you didn't consider. I think Linus's recent patch series is a good example. He saw a potential optimisation, did a proof of concept but missed a race condition. He's now handed it over to the person that normally works in that area and knows the details. Can the optimisation actually work? Who knows?
Weird, i disagree with Linus.
AI is not a tool. Its an addiction. Maybe the first dozen times (more likely: never) you use it "as a tool". Very quickly, you depend on it. You need it to refactor. Then you need it to debug. Then you need it to write. Next thing you know, you spend 0 time writing code (the fun part) and 100% of your time as a reviewer (the worst part). Suddenly it isnt fun. A few fools even get TWO AI MODELS, one to write, the other to debug.
Its a virus. I cannot tell you how many 300+ line commits ive rejected because they straight up dont do what the person said they do or also do some extremely asinine 2nd thing that breaks other pieces of code (which they would know if theyd actually used it).
A tool requires input. A hammer can never do anything i dont explicitly tell it to do. A good word editor does exactly what i want it to do. A good compiler does exactly what it says it will do. AI does not. Ai is a glorified autocomplete, and i dont want my code to be automatically generated. Its asking for thousands of issues down the line. If AI was any good, Windows would be a usable platform by now.
Dont get me wrong; i know that it is sometimes used as a tool by people, and meticulously checked over, but more often than not, it isnt.
I feel about ai like i feel about cars. 90% of the idiots out there drive really poorly, but that doesn't mean responsible people are smart to never drive a car. Same with AI, overwhelming majority of people are using AI incorrectly, i.e. recklessly dangerously, but that doesn't mean responsible people should never use it.
This is really an ignorant and arrogant take at AI.
Why do you say that?
AI is much more than the quality of the code. It's also about stealing the knowledge, overusing Earth's resources like water and power, pollution, killing consumer computers through RAM and memory shortage in general - the very same computers that run Linux -, making average people more ignorant, risking another gigantic bubble burst and so on. And as a cherry on the top, comparing AI to compilers. Riiiight.
Have you actually looked into the real environmental impact of AI is compared to, say, eating meat, driving a car, or taking a yearly plane flight? It's a rounding error. Have you looked into the numbers? Compared the environmental impact of your AI use to the environmental impact of taking a yearly vacation on a plane?
The RAM shortage is real, but very temporary. Check the news today! Chip stocks like Sk Hynix are way down because people are worried about over-production. The fact that a new technology caused a temporary shortage in a resource just isn't even a coherent argument against the technology. There's discussion to be had here about whether we should as a civilization reserve some resources for other industries. That's a discussion we could have. But the idea that AI caused a temporary shortage in RAM, so that makes it bad is... like, really ridiculous? You must see that that doesn't even make sense as an argument?
These are not arguments that hold up to scrutiny. You should consider whether you accepted these arguments because you looked into the details and they actually make sense, or if you accepted them because they're telling you what you want to believe.
What about those of us who don't eat meat or fly on a plane ever? Are we allowed to hate GenAI?
You're free to hate GenAI all you like as far as I'm concerned. Just be honest about your reasons, and don't pretend it's about the environment if you don't care enough to even look at the numbers.
Sure, but surely the data centers illegally running gas turbines in Memphis and Vineland, New Jersey are causing problems that wouldn't exist without the current hype cycle.
Or with moderately competent community involvement and regulation. Fortunately we're waking those up... kinda...
Sure.
Do you think that's a reasonable reason to hate gen AI?
There are absolutely issues around the rollout of AI. I haven't seen anyone disagree with that. That's just not a good argument for being opposed to the technology. That might be a reasonable reason to hate xAI or Microsoft (and I do!). It's definitely a reasonable reason to have concerns about how we're doing this. You should be concerned.
It's not a reasonable reason to hate the technology itself.
Basically you are saying that me buying and throwing a plastic bottle in to the nature is fine, since one bottle is nothing. Also I'm sure that (new) data centers run on few AAA batteries.
I didn't say that.
Listen, I've got nothing against you hating AI. There's lots of valid reasons to be concerned, especially with the implications for the future of jobs and work. The way corporations are adopting AI is legitimately concerning.
Just don't pretend it's about the environment when it's not. That isn't just an unconvincing argument, it hurts the environmental movement generally.
It's the plastic straws all over again. You understand that the stupid straws debate did the environment no favors right? Fighting AI as an environmental issue is much worse because the AI is coming, and you will lose that fight. If you're arguing for more controls about water and power use, I'm with you. If you're arguing for a carbon tax, I will sing your praises. If you're trying to fight the idea of AI itself, and pretending that it's about the environment, that's where I have a problem.
But Ed Zitron makes me feel fuzzy and warm when he name calls everyone and therefore he must be right
Man WAY2GO you really provided a wealth of information proving your point is more valuable than his!
I have mixed opinions on his stance towards AI.
I confess, I’m guilty. I’ve used AI to diagnose random issues with my little personal blog (please don’t hate me, I’m just a man without a degree). That said, I understood the simple code well enough to see what I was getting wrong and comprehend the AI generated suggestion, so it wasn’t all that different than looking things up on a wiki and implementing it. The primary difference, for me, was that I had already spent two weeks digging deeper and deeper for my answers, only finding outdated or incorrect information, and Claude delivered the correct answer in about 10 seconds.
In my opinion, the reason it was a good idea to implement the AI generated code was that I understood it. There was a human brain looking at that result and going “aha! I see what I was missing!” If I hadn’t understood what I was looking at, it could have been doing anything, and I wouldn’t have felt comfortable implementing it. If I hadn’t at least tried to the old fashioned way first, I wouldn’t have used it. But after weeks of hitting dead ends and with the knowledge that there was an extremely powerful tool that could give me my answer in seconds, it began to feel foolish to continue punishing myself with wasted time.
In Linus’ case, I think it is safe to say he understands enough about Linux to interpret the AI suggestions and be a safeguard against slop. I’m not too concerned about his contributions to the kernel if he’s using AI to think outside the box or diagnose a complicated problem.
But! There is a major risk factor when people start implementing code that they don’t understand, or when it is a crutch for their total lack of comprehension. Vibe coding and using a diagnostic tool are different things. I hope Linus gives some real thought to the realistic limits that need to be in place for people who aren’t knowledgeable enough to filter AI suggested code through their own brains, because it wouldn’t take long to destroy a reliable system like Linux if they incorporate AI generated code with no experienced human in the drivers seat.
Apologies to those who just fundamentally hate AI and think I’m a fool for using it, I just wanted to share my experience as an unschooled computer nerd who finally gave in and tried it out a couple of times.
Educated computer nerd here. AI has fundamentally changed how you do CS in industry, it's completely unavoidable at this point and not using it in some capacity is doing yourself a massive disservice. It's able to catch edge cases we'd never have caught otherwise. The important part is indeed understanding what it's doing and keeping contributions organized/concise.
That doesn't make it any less scary from an ethics and model oversight perspective, but from the perspective of the project it is fundamentally just another tool. Linus appears to feel this way too. Frankly I'd be more concerned if there weren't people using AI on Linux, especially to validate kernel level security.
I mean, I use it because I'm forced to, but I don't feel much in productivity gains (not using agents though). Sometimes it's very, very handy, but most of the time it's like talking to the stupidest rubber ducky, and the frustration it gives me is not worth that imo. Others might make a different calculation, but ultimately if search engines didn't enshitify themselves into obsoletion, and LLMs didn't produce the worst, most verbose, and sometimes objectively wrong documentation, I'd not see an upside at all. But here we are.
If it was in truth inevitable then there would be no need to constantly shriek "It is inevitable!"
There is also evidence that deferring critical thinking to an LLM causes cognitive atrophy.
https://arxiv.org/abs/2506.08872
Conversations on this subject will continue to stagnate and miss the important points as long as people continue to act like the discussion of ethics is the same as the discussion of efficacy. People who disagree with AI on a principle of its impacts being negative will often call it “a lying machine which is regularly wrong and offers no utility”.
May I suggest that people with criticisms of the ethics and impact of AI take a charitable position towards the utility offered by AI. Accept something along the lines of a positive account given by an experienced software engineer - it’s useful, it presents risks but also it presents new ways to get certain work done well in much less time etc.
Even if you do not believe this, the point is it is a much more ambiguous argument than the ethical argument or the concerns over the impact of AI on the species, environment or society. You weaken those arguments by putting them along side poorly justified, somewhat falsifiable claims that AI just doesn’t work.
I think he's kinda right in that competent engineers can write good code using LLMs. I don't know whether they're actually more productive that way though. The problem lies in the fact that incompetent engineers can write a lot of bad, but good-looking code using LLMs, which might be a pretty big burden on the maintainers.
Also, comparing LLMs to compilers doesn't really work in my opinion. Compilers follow strict rules and translate one kind of logic into a different kind of logic, while LLMs turn natural language and vague notions into some kind of logic. Writing code is telling the computer precisely what it should do and prompting is vaguely telling it what end result you want and those are two fundamentally different things.
the thing is - no other tool (currently ^yall_cant_read^) in our industry, and others, out there involves as much active destruction in both the working class and environment in its use.
that’s it. that’s the bit.
for those not understanding: the word “in” above implies in current use not that i am comparing AI’s effects with the invention of knitting machines or electricity.
No it's just the newest tool that does that.
Are your clothes machine knitted? Do you know you're destroying the jobs of people who hand knitted clothes? And let's not forget the art form of hand knitting that is now being insulted by people who buy machine knitted clothes without a care. And let's not forget about the environmental impact of knitting machines that require electricity while hand knitting doesn't!
The same could be said about so many things.
There isn't really a fundumental difference thar I can see between AI and other tools, at least that I can see.
GenAI lets teenagers nudify their classmates, creating child porn. Twitter even let you do it directly on the site earlier this year.
Students are using it to do their homework for them and learning nothing; at least with CliffsNotes you got the gist of what you were supposed to be reading. We shouldn't want a dumber society.
GenAI also makes it substantially easier to post propaganda and spam. Bots on the internet were already common but now trying to find genuine art or conversations is like pulling up steel from ships that sank before 1945 because everything since then has been irradiated.
The US's dumbfuck government almost started a war with China earlier this year thanks to someone using GenAI. And yeah everyone knows they're a bunch of fucking idiots but that doesn't change the fact that it wouldn't have happened if LLMs didn't exist.
Calling something just a tool is whitewashing who's making it and promoting it and what it's used for. Stop saying that AI is just a tool and it only matters how it is used
AI in school is an interesting issue, we can't discount the fact that there are benefits to using AI in school as well. It needs to be up to countries to regulate usage and educate children on it's use. Socialist countries seem to be doing a pretty good job at this.
"sources say" hmm I wonder what those sources are, I'm sure the CNN would never extrapolate information out of nowhere to spread sinophobia and call for AI regulation (which is exactly what AI companies want at the moment.)
I'm not sure how calling AI a tool is whitewashing. Of course tools don't exist outside of their material conditions, however, AI has many benefits, it's just that capitalist economies will always use tools for the detriment of workers. Socialist countries have economies that are orientated to benefit the working class. They recognise the problems that arise from AI and work to fix them while using its benefits as much as possible to liberate workers rather than burden them.
the newst tool happens to be the one doing the most destruction, so.. yes.
also: i do knit, lol. just a coincidence in example.
I don't think AI is the most destructive at all. AI is responsible for around 1% of carbon emissions (do your own research if you want a proper source). There are so many things, such as cars that emit way more than this.
As for job loss there are so many things that have eliminated so many more jobs. Tractors, knitting machines, production lines, automation in factories etc. No one says we should stop using tractors so we can have more jobs for farmers.
That's funny, I've got a half knitted scarf somewhere but I've never gotten into it properly.
Now do it for the planned capacity of every AI data center currently planned or under construction or waiting for compute manufacturing by the big tech monopolists, and the trillions of dollars of labor that they're telling the world they'll replace whether the product is capable of it or not.
The bot doesn't have to be able to do your job to displace your labor. It only has to convince your boss that it can. Right now the big tech monopolists are actively convincing the entire world's bosses that they can replace your labor. This doesn't have to be true to be a labor issue, and it certainly doesn't have to be true to be a devastating economic bubble.
No you. My statistic still stands, if you would like to provide a different one please do. Maybe it will be 2%.
Your second paragraph is shifting the blame away from capitalism and towards AI. Economic bubbles and mass layoffs are features of capitalism, this is a pattern that has been repeating since capitalism was born. If not AI it would be something else.
I don't believe they'll finish. If you do believe they'll finish and succeed in ripping communities off as planned, then I think you're also predicting that this stuff really works and soon.
get back to the scarf! i have a blanket im working on too lol.
but yeah. i think from our sw/eng industry, ai is at the least the one with the steepest destructive “curve”, i suppose.
ty for your reply.
Maybe one day, I would like a scarf.
Flying is almost surely more destructive, and pretty common part of how we work these days.
I think the real problem is that sometimes things like this come around that are disruptive to the workforce, faster than the workforce can handle the influx of new unemployment.
It will in theory sort itself out, but at the moment, were in a transition period.
That doesn't mean its bad, but it can mean that our governments aren't responding in ways that support the worker in the transition which is their fault, not the technologies fault.
Edit: essentially disruption will always keep happening, its how its handled thats the important part.
I do agree. We can look at countries with different economic and political systems to see which ones handle them best. Countries with socialist economic systems are able to use AI for the benefit of workers. While companies with capitalist economic systems use AI at the expense of workers.
I beg you people to just once honestly put the supposed resource use of AI into any kind of perspective. What do you think, how many prompts worth of AI did your choice to eat a beef burger instead of a bean burger cost the world? How many have you eaten, have all the AI haters eaten, across their life time?
You won't like the answer.
what makes you think i don’t use ai?
Could have said the same about the invention of electricity
The invention of the light bulb.
That has been SO much worse than the negatives we've seen with AI. If anyone is interested, that is a horrible story socially, industrially, environmentally, etc. Definitely worth reading about.
yes. indeed. two things can be true at once.
Well, the issue is that people want to turn off AI, but not these other things. It's like, only the newest bad thing counts. The bad things I like (like meat every day) are actually fine, somehow...
I was gonna say nice strawman and then I saw the instance and yea this dumb statement checks out 🙄
The cotton gin. The ink press. The player piano. Vhs. Photoshop. Audio in film.
You’re so wrong it’s laughable.
Explain to me how Photoshop is comparably destructive as AI is lol. Also the cotton gin replaces slave labour primarily. Idk if that's what you want to put up as an example here.
none of those are examples of things in our current era doing as much damage as ai lol. can you not comprehend the comment i made?
I’m no expert but a compiler does not equal and LLM built on stolen data
I don't have a personal or professional workflow that would benefit from AI. I have not been impressed with any generated art or writing (knowingly, so far). I pretty much only know it for the enshittification of search results. It might be great at coding, I'll never know because in my first 49 years of life I've only needed to code enough to get a D+ in my Pascal class. It might get great at generating art but I haven't had to generate art since my sophomore year of high school. I hear it's good at scanning lab test results which is great for future me when I need a lab test result but doesn't come up for me in the day-to-day. There are plenty of other useful things I'm sure it can do for other people but the stuff they say it can do for me right now is nothing I'm interested in having done for me.
It's not great at coding. I'm on about my fifth "rewrite large system that's a pile of AI slop" job this year, trust me.
Shitty devs will write shitty code using AI
Good devs will write good code faster with AI
I have for sure seen far worse hand written code than AI would produce
If it takes less time and energy to describe what you want in natural language than to just write the code, there are two possibilities:
In the end, only shitty devs benefit, while being prevented from becoming good devs, while good devs turn into mediocre devs.
"I want a method that inverts a binary tree"
Can you code that quicker than you can describe it in natural language?
No obviously not, AI could give that to you in ~30 seconds.
And it would probably be flawless, but you can review it yourself to double check
A heavily documented simple algorithm. Not an original, complex, reasoned out system.
This is really more a case of possibility #2. Specifically, why are you rolling your own in the first place? The whole reason the output of the AI is "probably flawless", is because this operation already has a million implementations. What's the value in conjuring up another one and adding it to the code that needs to be maintained? Yeah, binary trees aren't terribly complicated, but neither is
struct Vector3 { float x, y, z; };or any of its operations, until you realise you'll be running those operations an awfully lot, so it would be better if it had the right alignment and SIMD instructions were used in its operations, and of course, if you haveVector3you'll probably also need aMatrix4x3and aQuaternionetc., so in the end it's just better to use an existing solution.This is what I mean with turning good devs into mediocre ones: the worst solution becomes the easiest solution, so that's what people will lean towards.
Of course, there can also be situations like you have a very specific use-case and using something off-the-shelf would be sub-optimal, or you're working with old code and making their own was the decision the previous person or your dumbass past-self made, so that's just what you have to work with. But the AI is going to struggle more with those cases too.
Lastly, I also mentioned "energy". If you're trying to put in words things that aren't easily conveyed in English (or any other natural language) all the time, you burn through your cognitive endurance like crazy. This is why you don't want to be in a highly technical meeting in the morning, so being in one all day with a moron you can't really come to common understanding with (beyond the size of its context window) is even worse.
20 years of experience, for what it's worth, and a vehemently disagree. But to extend an olive branch, I think a big reason for the seemingly polar opposite reception of LLM coding online is the context.
If you give an LLM all the tools it needs to write, research and verify it's code, to test the result (for example, in a browser), give it access to all the context it needs, like documentation, use it for code that is new, but not largely novel (CRUD business app vs. invent a new algorithm) and a competent person uses it, it's genuinely great. Great in a way no sane person could deny.
Inversely, if you're working in a shitty environment already, without documentation, where everything exists as implicit knowledge in peoples heads, where you can't test and develop freely, where you're forced to use shit LLMs or "hardened" aka shitty tools, for a extremely uncommon domain, you might have worse experiences.
I completely disagree. I am completely confident that an Opus 5.5 will code circles around the average developer. I've also had to rewrite AI slop, but I did it with AI. Came out great. Turns out, knowing how to use a tool makes a real difference. It's like saying knifes are shit at cutting because a lot of dumb fucks keep holding it the wrong way.
Yes, the AI is not great at the task without intense supervision and direction. Hence, it's not good at the task.
Example I'm currently working on. Unqualified junior dev yelled at AI over and over to resolve a bug without looking at the code. The AI noticed there was a discrepency between "user_name", "username" and "host_name" fields in the database. The discrepancy applied to two users, resulting from a poorly thought out decision to use "user_name" for a new version of this app. Instead of recommending correcting the two records, it created an additional backwards compatibility layer to pull the entire schema of a table, itemize every possible combination of columns, and write a generic lookup system for every possible combination of columns. This was not even a production system that needed backwards compatibility, it was still in development (6 months of development by this guy, I'm his replacement and finished it in about 6 weeks). And it was not a generic system to deal with that table in general, this was only inside the context of a single auth API method. Without basic intuition, it completely failed to diagnose the problems of its own creation, and just kept endlessly expanding the system to account for its own mistakes that it forgot hours later. The codebase that resulted from this process is absolutely disgusting.
There's a reason for this. It's not designed to do symbolic reasoning. It's designed to passively infer symbolic reasoning from linguistic analysis. It doesn't know what the hell it's doing.
Human slop code existed way before we had auto-complete in our IDEs, way before Copilot start writing small functions and thus bad code is not because of LLM. AI slop is no different than Human slop, it uses a shit ton of ressource to generate it and is faster at getting a working output but bad codebase is not new.
Did they make you take Pascal in school, or did you choose it and then not like it?
That's pretty cool if it was required (or some form of programming), it is something I think everyone should at least try and be shown how it works.
My first couple of programming classes were in Pascal. Then there was the third class where the professors had written their own pseudo language that existed only on the chalkboard and in their own heads. At least I learned recursion from them. I had to teach myself every language I ever used on the job. I was EE instead of CS, so I suppose I shouldn't complain too hard, perhaps there were more practical classes deeper into the CS tree.
Multiple Pascal classes for EE thats interesting. You're not much older than I am and when we did Pascal my teacher said something like its not really used as much now but its a great learning language, but the 2nd class just moved onto C.
I had a friend doing EE and they had Perl as their programming class a few years later.
If I had to guess id say its Python now? That'd be practical as well.
I also leaned every language i used outside of school on my own.
Umm Spartan, what is your profession?
It works great as a fancy search engine. Google results have been atrocious for quite some time, and DDG isn't that much better. When you're looking for something hard to find, you may have to check dozens of results or try several searches. LLMs can do this and process the results much faster than you can, and can even give you the sources. Which you do need to check, because sometimes they still spout garbage.
Most other uses are crap unless it's used by an expert in the subject matter. But search can be useful for anyone. Note: this is not the AI results at the top of the search results; those are still crap. I'm talking about telling the LLM to do a thorough search with a lot more relevant context than the search usually has.
They still.kind of suck gibe you an answer and you say but blah. And itnfoes oh yes in version x.x they changed. I mean if unwanted old outdated in found use Google. Since looking for technical studfnis hard cause so much is out of date.
I got a stroke reading this.
They're bad at dates and age too. That's absolutely true. Like I said: always verify. They're good for finding stuff, not at being correct.
Genuine question. I'm not a developer, I didn't study computer science, and I'm a pretty basic Fedora user. But why does everyone harp on Linus Torvald's opinions like they're handed down from heaven? Windows still dominates the market, but no one cares what Bill Gates says. I love A Song of Ice and Fire but I wouldn't take writing tips from George R. R. Martin. Richard Stallman's opinions on GNU/Linux are memed endlessly. Linus is literally just a dude.
Bill Gates doesn't work on Windows anymore and hasn't got years.
George R R Martin doesn't publish books anymore and hasn't for years.
Richard Stallman is...a weird little guy.
Whether you like it or not Linus is still the Linux guy.
I feel like Linux having a guy is antithetical to Linux but I guess I'm just in the minority.
I know what you mean but he literally created it and still runs the show and it's named after him so like 🤷♂️
On that note, OpenBSD appears to be a likely refuge now that Linux has been abandoned to slop.
https://marc.info/?l=openbsd-tech&m=177411620801633&w=2
What market?
Desktops? Sure.
Servers? Definitely not.
Supercomputers? Like, none of them.
...
Phones? Nope.
IOT? Can't even run on modern laptops right.
You should add laptops to the top of your list...cause that's what dominates on productivity...and that's nearly ALL windows! What I think truly dominates is an offshoot of *nix...Android and IOS. Don't you just find it funny when you walk in a mall and see a jacked up Kiosk with a windows 10 taskbar sideways? ...Press any key to reboot!
When people say desktop laptop is implied in that too.
Gamers are about to step into this conversation....you sure you want to continue? :-)
Sure buddy, ever heard of chromebooks? Macbooks? Frameworks? Any serious dev that has a choice runs linux on theirs afaik
No...what's a chromebook? Did you not read my response? (hint: 2nd comment)
Asahi is cool but not ready for primetime dev work. Performance is ok, hardware support is spotty, battery life is appreciably worse. Any dev with a macbook is running MacOS 9.9/10 times
As someone who did study computer science and has worked in embedded processor space (so one of the people who's written assembly by hand he's talking about).
I've never understood it either. People take his word as law blindly and it's never made sense to me. Yes he has skill, but so do hundreds of others. Yes he's opinionated, that's certainly not a reason to hero worship him.
For ten+ years people have been pointing out he's basically a single point of failure for Linux and begging for a succession plan and he's been too stuck up his own ass to accept that he isn't the only one who cares about all this.
This has lead to some extreme cult following behind him for some reason.
I've translated assembler by hand and punched the hex codes into an EPROM programmer's keyboard. This would appear to make me qualified to say that he's not always right and, in most cases, fuck AI.
is Greg Kroah-Hartman not the succession plan?
As far as I know no. The plan is for the the top maintainers to gather and choose a successor. Though if you have more up to date information I'd love to see it. I could be behind the times on that.
Using an argument from authority to denounce someone in authority and their arguments? Cool. Might've worked if I hadn't seen you mention recently that we shouldn't use AI 'because Trump and his cronies do'.
Because he's not some guy on the internet saying his opinion, he is maintaining the most relevant Free Software kernel. So if he goes nuts, there is great potential for great disaster. I don't think there are Linux distributions that use forks of the Linux kernel. (What notable forks are there anyway?) At the moment, they all depend on the original Linux kernel. There is currently no alternative. It would take a lot of disruption and a tremendous effort to establish and switch to a fork of the Linux kernel that is free of Linus' influence. Today, I'd say jokingly I will switch to FreeDOS. But if this sloppification goes on, I guess I will have to.
https://codeberg.org/ethical-foss/open-slopware#other-operating-systems
FreeDOS isn't immune from AI either.
Because people's love of cults of personality is endless.
Add on top of that the "AI" cult that saw this as an amazing recruitment opportunity, and you get a nice combo.
That's basically the computer science equivalent of having a famous actor advertise a sect.
The difference is that Linus does not actively seem to be part of the cult itself rather than using it as the effective tool that it is for someone who already has his type of skill set. He can actually understand and double check inputs and outputs because he already built a foundation of great skill.
Most of the AI techbro community just wants an excuse to stop paying people anything for their labor and get eternal digital slaves to replace them. They aren't wanting to use it as a tool to help them unravel complexity, they are just using it as a smash and grab for power and control. Linus... isn't that, and maybe it's not great that he's embracing it now, during the height of controversy. I kind of wish he had waited for the crash an mania to die down a little and then show it as a useful tool for people who already have skills.
LLMs are useless for a lot of shit but they are pretty damn useful for code. I think some of his recent work finding bugs in old code is evidence of that.
https://nitter.miningtcup.me/twtayaan/status/2109668665951941101
I don't follow him at all, but I'd guess he's making a statement about it now precisely because of the mania and people constantly asking him about it.
Because he still the main maintainer and a developer on a piece of software core to a lot of modern computing.
If he is wrong on this it could ruin the credibility and reliability of Linux and cause a fork that could tear the community and ecosystem apart.
If he's right he's proved that AI can make maintainable, reliable core infrastructure code and the field will change to reflect that. Can't tell your boss AI sucks and is unreliable and I won't use it if AI code is working well at the base of modern computer software.
Why wouldn't you take writing tips from GRRM?
Because then I'd never finish my masterpiece. /hj But more seriously, a lot of his stylistic choices are ones I try to not replicate in my own writing. The extremely detailed descriptions of food, clothing, woman's bodies, etc. I love the story he tells but find the writing itself can be a chore to read sometimes, if that makes sense.
I don't write but I get what you mean by detail. Some books just go too verbose on the mundane.
Cause the man clearly can't finish his fucking book
Linux is his.
He gets to make the rules. We get to decide whether to bail.
I don't understand confusion or entitlement over this.
It's okay, none of the most hardline anti-AI folks on here are, but they certainly like to pretend they know the tools and the industry.
Most? Where's the data? This hard-line anti-LLM guy has a degree in computer science and has developed neural network systems for a previous company. Many on here that I've talked to about it fall under similar category.
Many? Where's the data?
Proving yet again AI bros have lost the ability to think. "many I've talked to" <- that literally is the data. I'm the data. It doesn't matter what the number is.
As opposed to your comment
Literally me existing already disproves your comment.
I have strong suspicions you don't actually exist.
Except that a compiler is not a stochastic machine.
here is the email this screenshot is an excerpt of; the specific "X" (the thing which the message Linus is replying to here suggested perhaps AI should not be used for) was the writing of commit messages.
That does change things a bit
I think if I had taken an icecream scoop to my forebrain I would have experienced more intellectual development than reading the contributions of the sloppers.
You all need to read about enclosure and learn to make arguments without analogy.
LLM is like a compiler? Then please replace the C sources with English and let it compile!
Looks Under The Generative AI Hood
Sees Snakeoil Salesman Selling Labotomies As The Future
Does Linus support steaming artwork and a fascist surveillance state? Because using and supporting and entrenching the technology use is exactly that.
The amount of bootlicking Linus has been doing recently is insane. I expected this bs from anyone but him. Such a sad reality we are living in.
I just want to say I was here before the downvoting started.
My understanding is that AI/LLMs can generate code that works, but often that code is more bloated/convoluted than code written by a decent coder, making it less efficient and more difficult to maintain. I can see that being acceptable for many in situations where hardware/performance/maintainability aren’t a concern like a one-off app or proof of concept. But for something as low level as a kernel I would think that performance and maintainability are the most important things since slowdown in that part of the system ripples up to affect everything a user does anywhere in the operating system.
I understand that a skilled coder can take AI output and clean it up into something at least approaching the quality of their own unassisted code, and I’ve been led to believe that Linus is able to do that for himself. But I fear that other/future maintainers of Linux won’t match that caliber, and it will start to feel that AI bloat. I think I’ve also read of Linus complaining about a torrent of low-quality, AI-based submissions and so I also fear that this will lead to him restricting who can participate in the work, overcompensate, and end up in a situation where Linux becomes less open and more closed.
So here's how I see it: There is a big difference between "AI generated" code and "AI assisted" code. Or just anything.
To put it in terms most people can more understand let's say you were writing a book. If you have an LLM generate the book its gonna suck and be slop. But if you write the rough draft, then paste it into an LLM and say, "Look through this and find typos and other mistakes and point them out to me". Then you go and check the things it finds and fix them yourself, that is fine. It's just a really good spellcheck at that point. Even if you say "suggest a fix" it's fine. For coding you could even tell it like "Take this line of code and suggest 5 different ways to do the same thing that might work better" then test them all out and see which one you like best.
I have personally used an LLM to help clean up my dotfiles for example. I don't want to read through all that shit and find the old junk I need to remove or change. So I just throw them at an LLM and have it point things out to me. Or say you know you have a file but don't know the exact filename. Go into the folder its in and dump the output of ls into an LLM. Tell it what file you need vaguely. Have it read them and give you the correct filename. Then check it yourself.
The problem comes from people who just generate stuff and don't do the work of double checking it themselves.
One thing people forget is that many human coders absolutely suck at what they do and or don't care.
80% of the code I've seen at clients is shit, some of it is hilariously bad, like absolute dog shit.
I've had a meeting where I had to explain the codebase to a new joiner, we were crying laughing at how bad the code was.
This particular fortune 500 company never cared to fix it, they just wanted more features.
So yes AI can be shit, but god I trust it more than some of the shit I've seen humans write.
Shit devs will make shit code with or without AI. Good devs will make good code, just faster with AI
There's nuance here though, LLM's actually do a better job of code that looks like it should work.
You get an unskilled junior in the mix, their code is obviously bad.
You get a semi decent LLM with a competent harness in there and you get a mix of code that works and code that nearly works.
There should be competent peer review either way, but the effort for each is potentially an order of magnitude different.
Then add on top that the LLM can churn them out way quicker than a junior could possibly achieve and you have a recipe for a huge review burden.
Yea using a junior or a agent both comes with issues and risks. You need to know what you're dealing with and shouldn't trust either blindly.
Ultimately someone has to take ownership of if shit goes wrong, and someone has to understand what the code does and why.
If you don't something has gone wrong.
This was more true last year but developments in both the frontier models and the harnesses used to control and direct those models have gotten better to the point where I'd say the code is better then most devs. People on here don't want to admit it but if you did a "coding turing test" between Claude and some random dev coding by hand I'd bet most of the time a dev would say the AI output is better.
People said this a year ago and I tested it then and it wasn’t great. I’ll believe it’s better but it’s hard to believe it’s “fixed”. I’m sure I’ll get a chance to try it again sometime. No way I’m paying for it myself lol.
Completely disagree with his take on this.
He doesn't get it... Have the big corpo-shilling libertarians rotted his brain?
Like, imagine if using "AI" consisted of whatever big corpo SaaS you subscribe to sending your prompt off to someone to work on then send the result back to you. Isn't that an idiotic dependency to have? And why the fuck even go through the middleman? Oh they have the "best workers?" Fuck that. Nvidia makes the best cards, but you don't see me buying from them.
Do people like Linus not realise how silly it is to offload your work to some middleman service that just steals work from people who do not consent to sharing their labour? Who are desperately throwing insane, GDP-level boatloads of cash at it this unknown technology in the hopes that they'll somehow get people hooked on their particular chatbot?
The intentions of these companies are not good. Their entire valuation is not based on being a "force multiplier" or whatever bullshit, but on being able to threaten or fire workers if they don't slot themselves into the newest enunciation of hype-based capital.
Anything you do to willingly boost or massage these corporate services is enabling actually vile companies to continue puffing air into their stupid pointless bubble until it bursts and the next generation has to eat tuna helper sans tuna just to be able to afford to live.
What is stupid is referring to AI as if it was just one tool when actually it's a bunch of tools, some of which are easier to use sensibly amd some it's near impossible to. Fancy auto complete like it's being compared to sure, fully automated do what the fuck you want agents, yeah sooner or later that is going to go wrong and no one is going to know where or why.
They might be fun and interesting, that doesn't make it a great idea to go all in in my opinion, certainly not for anything you would consider important.
Does he... think LLMs are deterministic?
Do you... think it matters? Tool generates code, you look at it, check it and correct it. You send it to a maintainer by giving your name and email, you are now expected to help maintain this code if it gets merged, it has your name on it.
Explain to me why writing the code by hand or with a "deterministic" alternative why would change the quality of what you do or are expected to do in future.
writing the code is the easy part, having a good mental model of what's going on under the hood and planning ahead is the real challenge. if a chatbot puked out the code it's easy to assume that it's right because it worked under your conditions but you might miss stuff that you'd otherwise acknowledge if you wrote the code.
using the chatbots directly discourages thinking and you're deskilling yourself in the long run, depriving yourself of hands-on experience that you'd usually pick up through your coding career.
having a chatbot generate code for you also usually goes hand-in-hand with writing more code faster, which tends to be bad if you like stable software and want to prevent maintainer fatigue.
and that's before we get into the ethical and legal problems.
Slop coders see Kernighan's Law and think "I can do so much worse"
He compared an LLM to a compiler suggesting that the concerns about using them are the same as suggesting that everyone should write in assembly.
The comparison is invalid, since one mode is deterministic, while the other is not.
LLM-driven development leads to programming via the abstraction of the model, disconnecting you from the source to the point where you don't understand what it's done or why. He's suggesting that this isn't a problem because an LLM is just like a compiler, which suggests that he's arguing in bad faith or doesn't understand the "tools" he's using.
You don’t need to check that the machine code your compiler made matched the code you wrote. You don’t need to check that the commit message an editor sent matched what you typed.
You absolutely do need to check and correct everything that comes out of AI because it’s at best like 80% correct, and that’s if the subject matter is easy.
Man, that's a shame
It's one thing to think AI is good. It's another to call others stupid for rejecting it
He's not calling people stupid for not using it, he's calling people that call for blanket bans on anyone submitting code that was made with AI stupid, since it should be judged off quality instead.
How do you parse the last part, if not calling people stupid for not using AI?
Man I've seen others try to twist those words the same way even though it's on the same level as one of those elementary reasoning problems you'd find on a basic IQ test (Jack caught an std after interacting with koalas - a) all koalas transmit stds, b) some koalas transmit stds, c) pandas are free from stds, d) Jack should spend more than $5 on his hookers).
So unless you actually are stupid and this is why you're choosing to (defensively) misinterpret his words by picking the equivalent of c) here, it's clearly wilful misrepresentation of a very reasonable statement - "Of the people I admire, most don't shy away from cutting edge sutff"
This would be true if we were reading a paper on formal logic and all people communicated exclusively through lean.
Why would he chose to include that section in this correspondence? Just as a non-sequiturs? Your analysis implies he had absolutely no rhetorical goal in including that section. Humans tend to have rhetorical goals when they write or speak so we must ask "what was Linus' goal" in that section.
Given the context, the fact that the email thread is in regards to letting LLM's make commit messages, and Linus being famously willing to insult the intelligence of others it would be foolish not to assume that "insult their intellect" was not a rhetorical goal.
Even considering rhetorical goals and context, without "I particularly dislike this guy", it's an enormous leap to twist those words that way.
Stop blaming the pandas.
So why did he include that part? No particular reason? Was he just making an unrelated remark about some of his friends? Accidentally included that bit but it was intended for a different email?
The simplest explanation is "similar to all the other times he's made back handed remarks or direct statements criticizing the intellect of who he was speaking to, he was criticizing the intellect of who he was speaking to".
This is an odd hill to die on. Linus is famously callous and rude in the Linux mailing list so to declare that this time he was just making a completely off-handed and unrelated remark about his bois is a difficult position to defend.
My last use of AI is in my post history and was fairly down voted because I was being foolishly ironic.
This said, I don't really disagree with his take here despite keeping away from AI as much as I am reasonably able. I do my own critical thinking and hate that I have to jump through hoops on the internet to have a simple question answered without some layman's AI summary that may or may not be even remotely correct or actionable. Even still, when I asked it to check on local real estate laws and provide summaries of possible actions based on a very strict set of parameters, it gave me information that my wife and I were able to bring to a human agent and have things work out better than we had actually expected.
I think a better way of putting it - and I'm still gonna ruffle feathers - is that AI is becoming the new electricity.
Put down your pitchfork and lemme explain.
When electricity came out, people rather reasonably pointed out that simply coming into contact with an electrical device could result in death if one wasn't careful. On top of that, the infrastructure required for it was expensive, unsightly, and at the time incredibly disruptive to daily life. It was going to ruin and disrupt entire industries that had existed for centuries because it could do things a human could not do unaided.
All of this is to one degree or another entirely true even through to modern times, but electricity has become a ubiquitous resource that humanity would now deeply struggle to live without. I won't dare compare AI that far as i genuinely believe humans would be significantly better off without it, but the initial Luddite response to electricity has a historical echo here.
I think Linus' point is more that it can undoubtedly be a useful tool when used in a measured, controlled, intelligent manner, and that it is a genie that is very clearly unwilling to be put back in its bottle. Like it or not it exists, and while we can absolutely be unhappy about it, being blind to its actual real world beneficial uses because the companies that make it are actual festering waste is a bit silly - especially when in the hands of the right chaotic neutral parties these companies are about to have a massive lesson in just how messy their tools have made things.
Very well put
Turns out my organic LLM occasionally vomits out sensible things, too.
The problem with the tools , it's not gonna help amateur or rookies to develop the skill without the tools, it seems make the illusion that it is capable to do so but it's not.
While for the veteran or those who know what they doing, of course it will produce various result , from diminish their entire person or just become another tool they wield for their work, depending on skill issue.
distant nuance bros' war drumming intensifying
Have fun with the incoming "contributions", Linus
why the highlights?
Except you can't with AI. It vomits a black box, and you have to decode it and test it before dumping it in a commit. Few do that.
Every once in a while I'll ask the DDG AI a question, and then try to confirm it with narrower searches. I think that's the extent of my use of LLMs.
And I feel guilty for it, wondering how much water and power it's taking and if the compute is poisoning the air of its human neighbors and keeping them up at night with a 70 dB hum.
There might be something to be said about using models locally, and using them as part of a human-guided process.
Note: I've stopped calling it AI in favor of LLM, partly due to the whole SI controversy in the US and partly in the hope that AI is still a larger category of learning systems that might lead to AGI, if we don't destroy ourselves first. Also the Jev model looks like an interesting development that indicates the ultimate tools might be other than a later model from OpenAI or Anthropic.
Note Also: I programmed in the 1990s and broke my brain trying to teach a computer to competently play a complex strategy game. I got pretty far before all my variable names leaked out of my head. These days, I write simple AHK scripts but that's about it. So I'm way out of my league when it comes to sound 2020s era coding methods.
Maybe the rule should be that if you use AI, just make sure Linus can’t tell you did.
I promise I won't submit anything to the Linux kernel that was generated by a compiler. AFAIK they want the code, not blobs.
I mean, it would be better if you could write it in assembler.
Fully agree with Linux.
Ai is a tool. Learn how to properly use it or not use it like you prefer, it's a goddamn tool.
If you are a bad coder you will still be a bad coder using AI. If you are a good coder, you will be a faster good coder with AI.
It's a fucking tool.
There is no such thing as just a tool.
Yes but unfortunately this specific tool is here to stay. Either we learn on how to use it properly and ethically and we regulate it, or we will be in deep troubles. Just denying and refusing to accept AI is here is dumb.
By regulate I mean at state and super-state level (not thinking of USA, but of the world) and define and accept proper ethical rules, both for individuals and corporations.
That blog post is just a bunch of insults to anyone not agreeing with the writer without actually giving any elaborate or truthful argument on why is not just a tool and it's always inherently "evil".
Some arguments made can be easily disproven. For instance the climate change one.
They main point of the post is that "no matter how you use LLMs". But then the argument given here it's highly dependant on how you use them. The amount of environmental damage of LLMs completely depends on which LLMs you use, how you use them and how often you use them. This just here it's showing the irrationally of the blog author.
But it goes one, next point is economics. Which is also 100% dependant on HOW YOU USE IT. Not all LLM are made by big evil corporations, and there's nothing in the technology preventing it being made by anyone.
For a blog post whose main argument is that saying "it depends how you use it" is wrong, have only give arguments that are dependant on how you use. The author makes broad ignorant assumptions about how people use LLMs, asserts that everyone uses them like that and then pretends to control everyone lives based of that. That way of thinking is highly dangerous and it's what have given birth to a lot of evil on our world. We need to be better, more rational, more open minded. Not falling for easy politics and actually do complex political thinking that won't give simple answers to complex questions.
Except Torvalds is now a bad coder with AI.
AI use demonstrably leads to cognitive decline. Looks like we have a great example in Torvalds, outside of the research papers already released.
These research papers talking about AI leading to cognitive decline seem so asinine when certain people's productivity has just skyrocketed under AI.
Sure the hand coding capabilities decline, but that's replaced by thinking around the actually product or architecture.
A huge sect of new Linux users and Selfhosters are talking about how much AI has enabled them yet we're still bringing up papers on cognitive decline that were written on models from 3-4 years ago when Claude Code didn't even exist.
Those papers, and the most recent ones, all point to use of LLM's as causing cognitive decline due to offloading problem solving.
Saying "they don't even take into account this way more capable model that makes me able to solve problems way faster" is like saying research into cancer causing materials is bunk because they don't take into account all the times I inject concentrated carcinogens directly into my testicles.
No one is saying using LLM's slow you down. They're saying that being slow was a feature not a bug and skipping it prevents long term learning. Is there a way to use LLM's that improve learning? Maybe. My worry is that even if there is there is little real inventive to do so.
Learning has just shifted, you no longer have to think of syntax and lower level things, but more the architecture of the code.
Many devs stop writing code and just become architects later in their career, their brains don't just wrot then because they "stop thinking"
I actually code way way more now than before AI, because it takes care of the shit I hate doing, and I can focus on the fun stuff
I could care less of what you think is asinine.
that is ... Debatable.
Yeah even writing lead to cognitive loss, because now people don't have to memorize everything.
It's just fear for the future and for change. And this is probably the most massive change in human history since fire was discovered.
But don't fear too much: the change is here and will stay, that's the only sure thing
Edit: what I mean is that accepting the fact AI is here and won't go away because we don't like it is the path to start regulating it at government level and agreeing on ethical rules to use it. Just saying you are a bad person of you use it it's like hiding under the sand. You lose.
source?
One example.
Got something that has more than 18 participants split into 3 groups for a total sample size of 6 people per control group? Not to mention its focus on AI assisted essay writing, measuring cognitive debt (not decline), using EEG to infer performance of higher cognitive function, etc.
No. AI is simply too recent and evolves too quickly to be demonstrably anything.
The studies you refer to are mostly flawed, and most importantly not about software engineering, where I assure you the problem solving part is not diminished, simply shifted. Don't let the memes that show people type "just do your best bro" into claude code influence you, they are just memes :)
Linus Torvalds is not an intelligent person. His argument style is extremely feeble, propped up solely from an ulterior, omnipresent appeal to authority. If he was just another poster, we would handwave his politics and social arguments as utter drivel.
He created some awesome software. Git, Linux. But the dude should have been kicked out of Linux dev a while ago. He is no longer required, and just runs his mouth constantly.
I wouldn't ve surprised if this was generated by Claude.
Why yes, let us compare corporate, proprietary, closed-source tools that inherently rely on environmental destruction, breaking licenses, and intellectual property theft in general, to basic, completely free and open source programming tools that have been in use for 50 years, and harmed nobody during their development.
Of course this guy does all his work on a MacBook. Because of course he does.
I've usually been with him, because he's very knowledgeable, and quite often his takes only seem ridiculous out of context.
But this is borderline corporate propaganda. If it continues, I'm afraid we'll all be moving to BSD.
He really thinks the tools you use do not shape and limit your results and yourself. I'd take a book carefully crafted in LaTeX anytime over slop pasted from Copilot into Word. What he's not even seeing is that AI is, above all, a source of input. No other tool, not the best LaTeX editor, takes care of my writing. AI is totally different. It is the author of Linus' code, it is his ghostwriter. He did not write it. He should not claim authorship. If he's not the author, he can't legally release it as Free software under a GNU license.
That’s not how this works. If he edits that code to his specifications, even using an llm, he’s the author.
This is how it works everywhere else and this is how it will work in this situation.
All right. And if I cp a jpg file, I'm a photographer. cp is my creative tool.
That's a dishonest example. The output of an LLM is never a copy of someone else's work. That's not how it works.
If you edit a photograph, then yes, it is your work, not as a photographer, but still your work. Even though you didn't edit the pixels one by one.
Id like to see Linus explain to me how a LLM works.
He is clearly ARROGANT and STUPID.
"Don't use a compiler then."
Does he not realize a LLM is not a deterministic system for "English to code". I've long be suspicious of his "genius" and this confirms it for me.
A giant gaping asshole riding off the work of the original unix philosophy and system, while spewing putrescence.
AI is just a tool, the user of a tool should be competent enough to use it properly
My calculator can do math for me, but I need to know how to use the calculator, the operations, rules in math, and basically, well, know math. My calculator is a good tool.
"IA" can "craft" a magic result without you needing to known anything about what the tool is helping you to accomplish. In a perfect world IA can be used to automate the most tedious part of any job (scaffolding, auto generated code for APIs, color grading, laso tool, etc) but in this world people use it as an excuse to not learn what they are trying to do. "IA" is a bad tool. Is not helping you, is doing it for you.
Never in human kind has existed a "tool" that does things for you without the need of any input or knowledge of the tool itself. "Engineer" comes from "who operates the machine", "IA" operates itself. I don't think there exist a real comparison with that "IA" does, but it's effects are clear, and are not good.
People taking shortcuts is not new.
30 years ago when I was in school, cheating was rampant. There's going to be about 15-20% of people who understand when they are cheating themselves, and the rest won't and never have.
I can think of plenty of tools that will do something for you even if you understand nothing about the tool, how it works, or the domains it is helfpul with.
I can plug a drill in a socket, put a random bit on and drill a hole without any knowledge of how a drill is supposed to work. But my hole won't be straight, will be too deep ir too shallow, too large or too small, and I might have drilled straight into a pipe. I don't understand how drilling works, I still drilled, but the outcome is shit.
I can mix a bunch of flour, eggs, sugar and chocolate in a bowl, then put it in an oven at max temperature on a random heat method and wait half a day. I'll have a bunch of coal, not a cake. I used an oven without any idra how an oven works, it did something for me. It's shit, but it did something for me.
Anyone can put the key in a car and start pressing random pedals to figure out how it works. Eventually anyone will get it... If you're not dead before then or you didn't total your car. Without understanding anything about a car, I did manage to move it and go somewhere. But I damaged some things in the way.
I could go on and on with as many tools as you want. But I doubt you'll keep reading.
LLMs in software development are the same: If you understand how to steer it, if you understand software engineering, it will have fantastic results. If you don't, it will have shit results. The only difference is that it is much harder to understand the insides of how an LLM model works, as that is another specialty. But just like you don't need to understand how an oven actually works to use it, you don't need to understand how an LLM actually works to use it.
Edit: Another great example is a camera. Anyone can take a picture, but without a good understanding of photography, most people's pictures will be shit. And the vast majority of users do not understand and do not hace the knowledge to understand how a camera works.
I agree. The main issue with AI in that respect is that there's no easy way to know how much it actually did vs how much value the user added while using it.
Things made mostly with AI are terrible in the long (or even short) run, but if important phases like analysis or design are mostly done without AI, the results are great, since it saves a lot of time for specific tasks.
I feel like this is the same as any tool, with the difference being that someone can fully hide that they used AI without it being initially evident, or lazy people use it as a shortcut to avoid learning. As long as you got all the base knowledge right and guide the AI yourself instead of letting it control you, it has potential.
Also, AI slop is so common nowadays that there's a bias towards them and all projects that use AI get labeled as bad by default, which doesn't help either.
You can know a hole is not straight. You can see it. Besides, if your house is built by someone that doesn't know how a drill works, well, it's going to crash down over you. Then why are we accepting works of people that can't and doesn't know how to properly use the tool?
Now, both of your examples have a person having to use the tool and know how to use it... IA is having the drill doing a hole by itself and just say "good enough". The house will crash down.
My point stand against both your examples. Both requires a human doing the work and the human to know how to use it to be useful. If the difference between a surgeon and a butcher is precision and knowledge, why are we allowing the butchers in on everything?
"IA" at best is a tool and at worst is a substitute. A "tool" that doesn't ask anything from you is not a tool.
I stand by it's not a tool. It's a substitute. But even by labeling as a tool, why should accept works from people that doesn't know how to use it and when? By following your examples we can't trust anyone claiming "I did this" with an "IA" until we or a professional inspect it, and since you can't inspect the absurd volume of things "IA" spits, the only solution left is ignore it all. Either you trust blindly on the guy that says he knows how to use the drill and risk your house falling down, or only accept works of a professional, and if you only take the job of professionals then what's the use for "IA" then?
When can we get an OS truly developed by AI? ...like ground up from Claude or OpenAI?
Might still be better than Windows 😅
Guess we'll just have to stick with spark for now.
Slop harder, Torvalds. You fucking lost it, and you're ruining Linux for everyone.
At least the other kernel maintainers still check things and catch problems, because clearly Linus Torvalds doesn't care anymore and would rather make false equivalence excuses.
So Linus does not respect me. Nice.
It’s mutual.
I've used Ai to try to make me operating systems and programs. All it does is give shells that you have to code the rest of what you want into it.
Then again, I don't pay for Ai and I've pretty much exhausted everything I wanted to ask chatgpt within the first 4 months of using it. Then I was able to make gemini think it was a different universe for a while until it caught on. Then I stopped using it as much.
It was fun for a while, Ai is still something you have to pick and choose and use good judgment with. It's not an end-all of coding.
Linus accepting AI now says more about how far the technology has come than anything else. AI today is not the same as AI a year ago.
I agree with Linus, because AI is just another Tool that won't go away. The Problem is not AI itself the problems are
Can't help but laugh at you guys correcting literally the undisputed #1 software maintainer in the world. AI is just a tool - if you fail then it's a skill issue.
He used to be cool, but then he started following some corporate hype trends that make me suspicious, I think he is nearing retirement and doesn't want to fight corpos anymore.
The moment he accepted Rust officially on the kernel was the time I knew he got a bit too old to care anymore.. This is just a continuation.
Also is so curious to see the venn diagram of people supporting Rust and AI being mostly a circle. My theory is that Rust takes every piece of joy out of programming and thus the idea of never programming again becomes so attractive if you are surrounded by Rust.
Even, reasonable opinions? Ooh boy that’s going to make some people irrationally angry.
BASED LINUS
01000100 01101111 01101110 01110100 00100000 01000000 00100000 01101101 01100101 00100000 01101001 01100110 00100000 01111001 01101111 01110101 00100000 01100011 01100001 01101110 01110100 00100000 01110010 01100101 01110011 01110000 01101111 01101110 01100100 00100000 01101001 01101110 00100000 01100010 01101001 01101110 01100001 01110010 01111001
::: spoiler 01110011 01110000 01101111 01101001 01101100 01100101 01110010 Dont @ me if you cant respond in binary :::
Piracy is good, and AI is piracy 😏
AI psychosis. Someone put him out of his misery. By replacing him I mean.
sigh. I think it's time to move to a BSD. Linus is gone.
Lmao your hero is here to call you out on your dogmatic Amish stance and you lot just outright reject him. Absolutely wild.
The Luddites weren’t opposed to technology. They were opposed to, in their words, “machinery harmful to commonality”.
It’s a shame that their name and cause has been totally contorted by the capital forces that opposed them, cuz it’s actually a pretty good lens for analyzing the societal impacts of AI.
After all, you can’t spell “machinery harmful to commonality” without AI.
Very interesting. Great knowledge, thank you. I guess I’ll have to revert to calling it being like the Amish.
Linus is talking about the worker using the technology. He likens it to other tools used in his trade. I don’t see him arguing about who should see the profits from the technology.
I think being dogmatic about strictly not using AI is not empowering the workers who would benefit from utilizing the tool in the correct context.
Luddite is correct. It just isn't the insult you thought it was. I'll add to the above commenter by saying that the Luddites were textile workers, which was the most highly skilled labor force in the world up to that point. They had I think a 6 year apprenticeship to learn to work with their very complex machines. The Luddites were the software engineers of their time.
They didn't want to do away with the machines, they wanted the workers to be in charge of what machines were used, when, where, and how. They were specifically concerned with the markedly lower quality of the fabric being produced with the new machines. In the burgeoning industrial capitalist economy, this could be considered an early example of enshitification. The Luddites were a labor movement concerned with the bosses ruining their craft with cost cutting measures, and they got branded as strict anti-technology primitivists so hard that it's still the common cultural understanding today.
There are of course no modern parallels for this and anyone skeptical of AI's usage today is definitely a primitivist that wants the AI genie put back in the bottle. /s
To quote Cory Doctorow "What's important isn't what a machine does, it's who it does it for and who it does it to."
Yeah that’s a great reply if you were trying to competitively comment but 99.9% of the online discourse against ai is completely against it. Including most of the discourse on this post.
Luddite is not the correct term. These folks are more dogmatic against Ai while Linus is literally talking about workers using it as a tool to make their work easier.
The work Linus is referring to is authoring commit messages in the Linux kernel.
He's arguing for letting the AI write the commit message. Honest question for you; if a person does not have the ability to articulate something like a change to the kernel, how do we know they understand the change?
Yea reactionary thinking is easier for most people than nuance.
He's a hero to a cult of tech bros. He's never been my hero and there's many in the Linux space who have been calling him out and asking for a democratic succession plan for ages. He's not some all knowing perfect being.
linux has a succession plan, though?
Sort of. The plan is "convene to make a plan" 72 hours after Linus's git stops updating. I should have been a little more clear though that people have been calling for a democratic secession plan to happen now instead of when Linus passes or steps away.
ah that does make more sense
Blasphemy!!
So not being thrilled about huge corpos throwing billions of dollars destroying the environment, ruining people's minds and jobs and potentially killing us all in 10 years is a "dogmatic Luddite stance". Got it.
You could say the same about the railroad bubble, when companies got over excited about rail.
Or the dot com bubble, the Internet destroyed many many jobs.
That’s weird, I didn’t see anyone approving of those things in the original posted text and I definitely didn’t imply any of that with my comment.
Unless I was mistaken this was about Linus saying that using ai is akin to using any other tool.
Actually with some thought, you’ve convinced me.
Hell the printing press is being used by our oppressors to kill the earth. I will be deleting my account and from here on out I will only communicate and read through hand written notes.
You dont know shit ^^
Can't take anyone seriously that uses the term luddite
I fixed it to correctly compare it to the Amish. Who are just against all use of technology beyond a point.
Linus is very obviously talking about the worker using a tool.
Y’all found a fake pedantic win and rushed me on it because winning is more important than conversations. Kudos on that.
Amish is also wrong. The Amish aren't against technology which is why you'll see Amish communities using things like credit card readers, industrial motors, and even computers.
The Ordnung, an Amish community's governing body, will review proposals for technological adoption every 6 months and hear arguments for and against adoption. The decision to adopt is screened against potential harm to the community based on a number of factors.
Honestly, the model the Amish use would be an excellent way to achieve a pro-social form of what the Luddites wanted to achieve.
Yeah? Your lust for winning an argument has you siding with the Amish. If they do things right abide by their rules and log off.
"Y-y-yeah well. If you like em' so much why don't you marry em'! Hehe got em'"
Please let's try and keep this above middle school. The whole point here is that comparing people who fundamentally disagree with this tech as "Luddites" or "Amish" betrays a lack of understanding of those two groups of people.
The Amish are modern day fools who reject this very thing we are on. I still say if you are so ardent to defend them you should abide by their rules and log off.
Your fake quote was a zing straight to my core. I’m not sure I will ever recover.
We will not agree here.
I didn't know Luddites were an actual thing back in the day, so i guess i learned something today, but according to wikipedia these days it's used as a term for people who are against new technologies in general, which isn't accurate cause i'm not against new technologies in general, i just have an issue with genAI in particular. I don't see how Amish would apply here either. Amish don't have a problem with specifically genAI, they just reject most things from modern society, including a lot of technology.
He's not a hero. He's an old man past his prime.
Based.
Is this the Linus Tech Tips guy? Isn’t he like a known asshole these days or something? Or am I thinking of another Linus?